🔗

Mappings

MONDO

MONDO:0012885 SRD5A3-congenital disorder of glycosylation

👪

Inheritance

1

Autosomal recessive inheritance HP:0000007

Affected individuals carry biallelic SRD5A3 variants. Most reported patients are homozygous for nonsense or frameshift alleles and come from consanguineous families; compound heterozygosity including a missense allele is less common and was among the first observed in adults with a milder cognitive phenotype.

Autosomal recessive inheritance

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"The majority of individuals with SRD5A3-CDG reported in the literature have homozygote or compound heterozygote nonsense or frame shift mutations, most with documented consanguinity"

States the biallelic requirement and the predominance of homozygous loss-of-function alleles arising through consanguinity.

◈

Mechanistic Hypotheses

1

Raised Polyprenol-to-Dolichol Ratio Model

polyprenol_competition_model CANONICAL

Evidence balance 2 support

The operative lesion is not the absence of dolichol but the accumulation of its unreduced precursor. Fibroblasts from a patient homozygous for an exon 1 stop codon, with no functional polyprenol reductase, contained normal amounts of dolichol together with elevated polyprenol, yet still hypoglycosylated transferrin. On this model the excess polyprenol competes with dolichol at the initiation of N-glycan assembly without being able to support glycosylation, so the pathogenic quantity is the ratio rather than the dolichol level. It also accounts for why complete dolichol loss is not seen: an alternative dolichol biosynthetic route persists, and its absence would be expected to be lethal.

Recorded as CANONICAL because it is the only mechanism proposed in this literature and is consistent with the fibroblast measurements, not because it has been tested against an alternative. The competition step itself is inferred from the lipid ratio; no assay of polyprenol-linked glycan initiation is cited.

Pathograph links

Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio -> Impaired Dolichol-Linked Oligosaccharide Assembly DIRECT Competition by the accumulated precursor at the initiation step is the proposed route from the lipid imbalance to defective oligosaccharide assembly.

Show evidence (2 references)

PMID:22304929 SUPPORT In Vitro

"Quantification of dolichol and unreduced polyprenol in the patient's fibroblasts demonstrated a high polyprenol/dolichol ratio with normal amounts of dolichol, indicating that high polyprenol levels might compete with dolichol for the initiation of N-glycan assembly but without supporting normal..."

The measurement and the interpretation this hypothesis is built on, stated by the authors who made it.

PMID:27480077 SUPPORT Human Clinical

"hypothesized that one explanation may be not the absence of dolichol (which would be lethal) but the increased ratio of polyprenol to dolichol leading to the phenotype associated with SRD5A3-CDG"

An independent restatement of the model, including the argument that dolichol absence would not be survivable.

?

Discussions and Knowledge Gaps

5

Why do the cerebellum and optic pathway bear the brunt of a defect in a biosynthetic step that is required by every cell?

KNOWLEDGE GAP OPEN tissue_selectivity_gap

Attached to

pathophysiology#Cerebellar and Optic Pathway Vulnerability pathophysiology#Systemic Protein Hypoglycosylation

Serum glycoproteomics shows hypoglycosylation is broad, affecting most measured glycopeptides, yet the clinical phenotype is dominated by two tissues. One hypothesis has been offered - particular sensitivity of the optic nerve to accumulated polyprenol, supported by the contrast with DOLK-CDG, which sits one step downstream, shares the skin and seizure phenotype and lacks the eye involvement. That contrast is suggestive but is a comparison between two small case literatures, not an experiment. Nothing in the cited work measures polyprenol in neural or retinal tissue, so the edge from hypoglycosylation to tissue selectivity is left with unknown intermediates in this pathograph. The obvious rival explanation has been checked and does not work. If affected tissues were simply the ones that need the most SRD5A3, baseline expression would track the phenotype; across twelve CDG genes in healthy human tissue it does not. What that survey does find for SRD5A3 is among the strongest tissue-specific departures from balanced biallelic expression of any gene it examined, which is a different kind of candidate and an untested one. So the gap is not that nobody has looked - it is that the cheapest explanation has been ruled out and the remaining ones are unmeasured in the tissues that matter.

Show evidence (3 references)

PMID:27480077 SUPPORT Human Clinical

"The optic nerve may be more sensitive to accumulation of polyprenol and this could explain why optic nerve atrophy/hypoplasia are ubiquitous in SRD5A3-CDG but not as frequent in other N-linked CDG."

The single proposed explanation for the selectivity, stated as a hypothesis by its authors.

PMID:41732066 SUPPORT Human Clinical

"with 245 of 291 altered glycopeptides decreased in SRD5A3-CDG"

Establishes that the biochemical defect is broad, which is what makes the narrow clinical selectivity require an explanation.

PMID:42472049 SUPPORT Computational

"the molecular basis for their tissue-specific manifestations remains poorly understood"

Confirms that this gap is recognised across the CDGs rather than being specific to how this entry is curated.

Why is there still no isolated Srd5a3 mouse model of this disease, and does the one published mouse deletion covering the locus tell us anything about it?

KNOWLEDGE GAP OPEN no_mouse_model

Attached to

pathophysiology#SRD5A3 Polyprenol Reductase Deficiency

The absence is documentable rather than merely apparent, which is unusual for a negative. A spontaneous mouse mutant carrying a megabase-scale 5qC3.3 deletion does remove Srd5a3, and homozygotes die at peri-implantation - but that lethality was traced to Exoc1, one of the other eight genes in the interval, and confirmed by making an Exoc1 knockout separately and by showing that a CRISPR-engineered deletion of the same interval lacking the Exoc1 defect does not kill the embryo. So the only mouse in this literature that lacks Srd5a3 tells us nothing about Srd5a3, and it would be a mistake to read peri-implantation lethality as this gene's phenotype. The gap that leaves is consequential: the disorder's defining features are ocular malformation and cerebellar hypoplasia, and neither the patient fibroblasts nor the C. elegans model can address them. Every tissue-selectivity question in this entry is unanswerable without a vertebrate model.

Show evidence (2 references)

PMID:26346620 SUPPORT Model Organism

"Genetic investigation revealed deletion of a > 1.2-Mb genomic region containing nine genes (Kit, Kdr, Srd5a3, Tmeme165, Clock, Pdcl2, Nmu, Exoc1, and Cep135)."

Establishes that the only published mouse lacking Srd5a3 lacks eight other genes too.

PMID:26346620 REFUTE Model Organism

"We concluded that peri-implantation lethality in Kit(WS/WS) was caused by a monogenic defect of Exoc1."

Graded REFUTE because it contradicts the reading this mouse would otherwise invite - that deleting Srd5a3 is embryonic lethal in mouse. The lethality is Exoc1's.

What is the alternative route to dolichol that maintains near-normal dolichol levels in cells with no functional polyprenol reductase, and does variation in it contribute to the phenotypic variability between siblings with identical alleles?

OPEN QUESTION OPEN alternative_dolichol_pathway

Attached to

pathophysiology#Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio mechanistic_hypotheses#polyprenol_competition_model

Two independent groups found residual or normal dolichol in cells lacking the enzyme and concluded that another biosynthetic route must exist; neither identified it. This is not a loose end but the load-bearing assumption of the ratio model, since that model requires dolichol to be present in order for polyprenol to compete with it. It also offers a candidate explanation for the within-family variability seen on identical homozygous alleles, if tissues differ in how far they rely on the alternative route. One candidate can now be excluded. Dolichol biosynthesis is no longer thought to be the single step SRD5A3 was originally assigned: DHRSX, found in CDG patients after SRD5A3 was, defines a three-step detour pathway that has since been shown to be conserved in yeast. That makes DHRSX the obvious suspect for the compensating route. It is not: in the yeast system, deleting the DHRSX counterpart produces glycosylation defects, low dolichol and polyprenol accumulation - the same picture as this disease - and expressing DHRSX rescues it while expressing SRD5A3 does not. The two enzymes are non-redundant steps in one pathway, so DHRSX cannot substitute for a missing SRD5A3. The question stands, with one fewer answer available.

Show evidence (4 references)

PMID:20637498 SUPPORT In Vitro

"The presence of residual dolichol in cells depleted for this enzyme suggests the existence of an unexpected alternative pathway for dolichol de novo biosynthesis."

The original observation and its interpretation, from the paper that identified the gene.

PMID:27480077 SUPPORT Human Clinical

"Different cell lines may also variably rely on the de novo pathway versus potential alternate pathways or salvage from turnover."

Raises the tissue-variable reliance that this question asks about.

PMID:42201967 SUPPORT In Vitro

"Deletion of TDA5 caused glycosylation defects, reduced dolichol levels, and accumulated polyprenol."

Shows that losing the DHRSX step reproduces this disease's biochemistry in yeast, which is what makes it a plausible candidate for the compensating route before the rescue experiment rules it out.

+ 1 more reference

Is there a corticospinal or neuromuscular component to SRD5A3-CDG beyond the cerebellar and ocular involvement that defines it?

KNOWLEDGE GAP OPEN corticospinal_involvement

Attached to

pathophysiology#Cerebellar and Optic Pathway Vulnerability genetic#SRD5A3 pathogenic variants

One patient has been reported with a proximal limb-girdle pattern of weakness, diffusely brisk reflexes and bilateral extensor plantar responses alongside the usual retinal dystrophy, ataxia and ichthyosiform skin change, and the authors propose this as an expansion of the neurological phenotype. This entry records the allele but does not curate limb-girdle weakness, hyperreflexia or extensor plantar responses as phenotypes. That is a deliberate line rather than an omission. It is a single patient; upper and lower motor neuron signs in a child with a multisystem disorder and marked hypotonia have several possible readings; and the features that would distinguish a genuine corticospinal component from an incidental one - imaging of the tracts, nerve conduction, muscle histology - are not reported. Curating three new phenotypes off one case would put them on the same footing as findings established across cohorts. Resolving it needs targeted neurological examination in an existing SRD5A3-CDG cohort, not another case report.

Show evidence (2 references)

PMID:41667393 SUPPORT Human Clinical

"neurological assessment revealed proximal limb-girdle pattern of weakness, hyperactive reflexes, extensor plantar responses with evidence of cerebellar dysfunction"

The findings themselves, in the one patient they are reported in.

PMID:41667393 SUPPORT Human Clinical

"We are expanding the neurophenotypic spectrum by reporting proximal limb-girdle pattern of weakness combined with diffusely brisk reflexes and bilateral extensor plantar responses suggestive of corticospinal or neuromuscular axis involvement"

The authors' own framing of the claim as a proposed spectrum expansion, which is what this discussion holds open rather than adopting.

Could supplying dolichol itself bypass the enzymatic block, and is there any experimental support for that outside human cells?

OPEN QUESTION OPEN dolichol_supplementation

Attached to

pathophysiology#SRD5A3 Polyprenol Reductase Deficiency treatments#Atorvastatin (experimental, not established therapy)

The one therapeutic candidate this entry carries, atorvastatin, works upstream: it was found in a motility screen and acts on the isoprenoid supply, not on the missing reduction step. A more direct idea is to replace the product. There is no human evidence for it, but there is a plant result that bears on whether the logic holds at all: in Arabidopsis, the orthologue of SRD5A3 is essential, its loss impairs protein glycosylation, and the resulting dolichol shortage is partially corrected by feeding dolichol. Two things stop that from being a treatment proposal. The rescue is partial and in a plant, whose dolichol requirements and uptake are not those of a human; and this entry's own mechanism argues the target may be wrong, since patient cells have normal dolichol already and the proposed lesion is the polyprenol-to-dolichol ratio. Adding dolichol would lower that ratio, so the idea is not incoherent - but it is untested, and the paper proposing it does so as a general suggestion for glycosylation disorders rather than for this one.

Show evidence (2 references)

PMID:26628744 SUPPORT Other

"Shortage of dolichol in PPRD2-deficient cells is partially rescued by PPRD1 overexpression or by supplementation with dolichol."

The rescue result the idea rests on. Graded OTHER because the system is a plant, which is neither a model organism in the vertebrate sense used elsewhere in this entry nor an in vitro human system.

PMID:26628744 SUPPORT Other

"Impaired protein glycosylation seems to be the major factor underlying these defects"

Establishes that the plant orthologue's loss acts through the same downstream consequence as the human disease, which is what makes the rescue relevant at all.

⚙

Pathophysiology

6

SRD5A3 Polyprenol Reductase Deficiency

Biallelic SRD5A3 variants abolish or reduce polyprenol reductase, the enzyme that reduces the alpha-isoprene unit of polyprenol to yield dolichol. Reported alleles are predominantly nonsense and frameshift changes; a missense allele modelled onto the predicted protein structure falls in a potential active site and is predicted to reduce catalytic efficiency rather than abolish it, which fits the milder phenotype seen in compound heterozygotes. What that enzymatic step is has been revised since SRD5A3 was identified. Finding the gene, together with its yeast counterpart, established the view that dolichol is made from polyprenol in one step; the later discovery of DHRSX in CDG patients replaced that with a three-step detour pathway, now shown to be conserved in yeast. The revision matters for this entry in a specific way: the yeast DHRSX counterpart cannot be substituted for by SRD5A3, so the two are separate, non-redundant steps rather than alternative routes to the same product. Whatever preserves dolichol levels in SRD5A3-null patient fibroblasts, it is not DHRSX standing in for the missing enzyme.

SRD5A3 hgnc:25812

polyprenol reductase activity GO:0102389 ↓ DECREASED

Animal models

C. elegans SRD5A3 W19X model Caenorhabditis elegans recapitulates Carries a patient allele and reproduces disease-relevant phenotypes at the organismal level, including a mevalonate-pathway readout that connects the lesion to the isoprenoid supply upstream of polyprenol.

Downstream

Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio Loss of the reductase leaves its substrate unconsumed, which is what raises the polyprenol-to-dolichol ratio.

Show evidence (5 references)

PMID:20637498 SUPPORT Human Clinical

"We describe a new type of CDG caused by mutations in the steroid 5alpha-reductase type 3 (SRD5A3) gene."

The report establishing SRD5A3 as the causal gene for this CDG.

PMID:20637498 SUPPORT In Vitro

"We found that SRD5A3 is necessary for the reduction of the alpha-isoprene unit of polyprenols to form dolichols, required for synthesis of dolichol-linked monosaccharides, and the oligosaccharide precursor used for N-glycosylation."

Identifies the exact enzymatic step lost, and why losing it reaches N-glycosylation.

PMID:33403770 SUPPORT Computational

"predicted that the p.(Ser154Pro) variant is located in a potential active site and is capable of reducing its catalytic efficiency"

Structural modelling of a missense allele. This is an in silico prediction, not a measured enzyme activity, and is graded accordingly.

+ 2 more references

Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio

Unreduced polyprenol accumulates while dolichol levels remain near normal, because an alternative dolichol biosynthetic route persists. The resulting shift in the ratio, rather than a shortage of dolichol, is the proposed proximate cause of hypoglycosylation: excess polyprenol is thought to compete with dolichol at the initiation of glycan assembly without being able to carry the glycan.

Raised Polyprenol-to-Dolichol Ratio Model

Downstream

Impaired Dolichol-Linked Oligosaccharide Assembly Competition by the accumulated precursor at the initiation step is the proposed route from the lipid imbalance to defective oligosaccharide assembly.

Show evidence (2 references)

PMID:20637498 SUPPORT In Vitro

"The presence of residual dolichol in cells depleted for this enzyme suggests the existence of an unexpected alternative pathway for dolichol de novo biosynthesis."

Explains why dolichol is not simply absent, which is what forces the mechanism onto the ratio rather than onto depletion.

PMID:27480077 SUPPORT In Vitro

"SRD5A3 is not the sole producer of dolichol since evaluation of fibroblasts from a child with a homozygous stop codon in exon one at p.Trp19X in SRD5A3 found normal levels of dolichol but elevated polyprenol levels"

Independent statement that a complete null allele still leaves normal dolichol, with elevated polyprenol.

Impaired Dolichol-Linked Oligosaccharide Assembly

Assembly of the lipid-linked oligosaccharide precursor on dolichol is compromised. Because dolichol also serves O-mannosylation, C-mannosylation and GPI anchor synthesis, the defect is not confined to the N-linked pathway, which is part of why the phenotype spans so many organ systems.

dolichol-linked oligosaccharide biosynthetic process GO:0006488 ↓ DECREASED

Downstream

Systemic Protein Hypoglycosylation A defective precursor is transferred to fewer nascent glycoproteins, which is what hypoglycosylation consists of.

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"This is a major pathway for dolichol biosynthesis for N-glycosylation, O-mannosylation, C-mannosylation, and GPI anchor synthesis."

Establishes that the affected carrier serves four glycosylation pathways, not only the N-linked one.

PMID:22304929 SUPPORT Human Clinical

"an enzyme catalyzing the final step of the biosynthesis of dolichol, which is required for the assembly of the glycans needed for N-glycosylation"

Places the enzymatic block immediately upstream of glycan assembly.

Systemic Protein Hypoglycosylation

Secreted and membrane glycoproteins carry fewer or truncated N-glycans. Serum glycoproteomics in patients shows the change is broad rather than selective, and the affected proteins include several whose deficiency has recognised clinical consequences - haptoglobin, plasma serine protease inhibitor, alpha-1-B glycoprotein, alpha-2-macroglobulin and ceruloplasmin. Hypoglycosylation is partial rather than complete: a patient with no functional enzyme still glycosylated most of their transferrin correctly.

protein N-linked glycosylation GO:0006487 ↓ DECREASED

Downstream

Cerebellar and Optic Pathway Vulnerability The cerebellum and optic tract are disproportionately affected in this CDG. The intervening steps between generalised hypoglycosylation and that selectivity are not established, which is why the link is marked as having unknown intermediates.

Abnormal serum transferrin glycoform pattern The transferrin glycoform profile is the direct laboratory readout of the hypoglycosylation, which is what makes it the screening test for this group of disorders.

Progressive Ocular and Skeletal Degeneration The adult-onset features arise on the same biochemical background, but nothing in the cited literature explains why they are progressive rather than congenital, or why the retina and spine in particular. Drawn with unknown intermediates rather than left as a second root, because a second root would imply an independent cause.

Ichthyosis Ichthyosiform skin change occurs in SRD5A3-CDG and also in DOLK-CDG, the disorder of the immediately adjacent pathway step, which suggests it belongs to the shared glycosylation defect rather than to polyprenol accumulation specifically. The route from hypoglycosylation to the epidermal barrier is not established here.

Palmoplantar keratoderma Part of the same dermatological phenotype as the ichthyosiform change, and attributed the same way.

Reduced protein C activity Coagulation regulatory proteins are N-glycosylated, and reduced activity of protein C and antithrombin III is documented in patients.

Reduced antithrombin III activity The same route as for protein C; both anticoagulant proteins were reduced in the reported siblings.

Show evidence (1 reference)

PMID:22304929 SUPPORT Human Clinical

"about 70% of transferrin (Tf) was correctly glycosylated"

Establishes that hypoglycosylation is partial even with a complete enzyme null, which constrains any model of the mechanism.

Cerebellar and Optic Pathway Vulnerability

The cerebellum and the optic pathway are the tissues most consistently affected. Nearly every reported patient has cognitive delay, an optic nerve abnormality with nystagmus, and ataxia, frequently with cerebellar vermis hypoplasia. The optic involvement is not purely a developmental malformation: it also progresses, implicating long-term maintenance of the nerve as well as its formation. One proposed explanation for the selectivity is particular sensitivity of the optic nerve to polyprenol accumulation, which would distinguish this CDG from N-glycosylation defects further downstream.

Downstream

Intellectual disability Cognitive delay is among the near-universal features.

Ataxia Cerebellar dysfunction manifests as ataxia, often with documented vermis hypoplasia.

Optic atrophy Optic nerve degeneration, distinct from the developmental hypoplasia and progressive over time.

Optic nerve hypoplasia The developmental, as opposed to degenerative, form of the optic nerve involvement. The source groups the two and describes the combination as ubiquitous.

Iris coloboma A structural malformation of the developing eye, reported alongside optic nerve coloboma. It is a closure defect rather than a degenerative change, so it belongs to the developmental half of the ocular phenotype.

Cerebellar vermis hypoplasia The structural correlate of the ataxia, present in a proportion of patients.

Hypotonia Muscular hypotonia accompanies the cerebellar and wider neurological involvement.

Microcephaly Reduced head growth in a subset of patients, part of the neurological phenotype.

Seizure Epilepsy occurs in a subset, in some from early infancy. As with the ichthyosis, the same feature occurs in DOLK-CDG one step downstream, so it is more likely a consequence of hypoglycosylation than of polyprenol accumulation.

Nystagmus Nystagmus accompanies the optic pathway involvement and is often the presenting sign in infancy.

Show evidence (4 references)

PMID:20637498 SUPPORT Human Clinical

"Patients have mental retardation and ophthalmologic and cerebellar defects."

The original description of the neuro-ophthalmological core, in the paper's own terminology.

PMID:27480077 SUPPORT INDIRECT Human Clinical

"The optic nerve may be more sensitive to accumulation of polyprenol and this could explain why optic nerve atrophy/hypoplasia are ubiquitous in SRD5A3-CDG but not as frequent in other N-linked CDG."

The proposed explanation for the tissue selectivity, offered by the authors as a hypothesis ("may be", "could explain") rather than a finding.

PMID:42472049 REFUTE Computational

"Tissues frequently affected in the corresponding disorders did not consistently display the highest baseline gene expression, underscoring that higher gene expression alone is a poor indicator of tissue susceptibility"

Cited as REFUTE against the simplest explanation for why this node exists - that the cerebellum and optic pathway are hit because they express SRD5A3 most. Across twelve CDG genes in healthy tissue that does not hold. It leaves the selectivity unexplained rather than explaining it, which is the point: the hypothesis above, that the optic nerve is unusually sensitive to accumulated polyprenol, is not competing with a well-supported alternative.

+ 1 more reference

Progressive Ocular and Skeletal Degeneration

A second arm of the phenotype: features that worsen with time rather than being fixed at birth. Adults with SRD5A3-CDG develop cataracts, retinal degeneration and kyphosis that were not present in childhood, so the disorder is progressive rather than purely static. This is the basis for the recommendation that children diagnosed with SRD5A3-CDG be monitored for these complications as they age. The retinal component needs a distinction the sources do not always make. It is progressive, but it is not late-onset: cohorts ascertained through eye clinics find visual loss by age three and a rod-cone dystrophy on imaging in patients whose retinas were never examined that way before. What arrives late is the diagnosis, not the disease. Cataract and kyphosis, by contrast, are documented as genuinely absent in childhood and present in the same patients as adults.

Downstream

Cataract Lens opacity developing after childhood in reported adult patients.

Retinal dystrophy Progressive rod-cone degeneration. Reported from the second decade in metabolically ascertained patients and from infancy in ophthalmically ascertained ones; the edge is kept on this node because the degeneration itself progresses, not because it starts late.

Nyctalopia Night blindness, the symptomatic counterpart of the rod dysfunction found on electrodiagnostic testing.

Myopia Short sight, reported alongside the retinal dystrophy and in some patients marked.

Kyphosis Progressive spinal deformity in older patients.

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"All individuals diagnosed with SRD5A3-CDG as children will need close monitoring for the development of cataracts, retinitis pigmentosa, and kyphosis as they age."

The clinical consequence of this node being progressive rather than static.

⬡

Pathograph

Use the checkboxes to hide or show graph categories. Hover nodes for evidence and cross-linked metadata.

●

Phenotypes

20

Eye 6

Nystagmus VERY_FREQUENT HP:0000639

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"nearly every affected individual reported to date has cognitive delays, nystagmus/optic atrophy/optic hypoplasia, and ataxia"

Nystagmus among the near-universal features.

PMID:22304929 SUPPORT Human Clinical

"The clinical features were psychomotor retardation, pathological nystagmus, slight muscular hypotonia and microcephaly."

Nystagmus in the index patient of an independent report.

Optic atrophy VERY_FREQUENT HP:0000648

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"why optic nerve atrophy/hypoplasia are ubiquitous in SRD5A3-CDG"

The authors describe the optic nerve findings as ubiquitous in this disorder, which is the basis for the band.

Cataract OCCASIONAL HP:0000518

Onset: ADULT

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"all have severe intellectual disability, cataracts beginning at 17 years and kyphosis at 8 years old"

Documents the age of cataract onset in a reported adult sibship, which is the basis for the onset category.

PMID:20700148 SUPPORT Human Clinical

"a novel syndrome consisting of mental retardation, coloboma, cataract and kyphosis"

Cataract is one of the four features of the syndrome that was described independently as Kahrizi syndrome before the gene was known, and that this paper showed to be allelic with SRD5A3-CDG. It is the reason Kahrizi syndrome is carried as a synonym of this entry.

Retinal dystrophy FREQUENT HP:0000556

Onset: CHILDHOOD

Show evidence (4 references)

PMID:28253385 SUPPORT Human Clinical

"Mutations in the SRD5A3 gene may cause early-onset retinal dystrophy, a previously underdescribed feature of the SRD5A3-CDG disorder that is progressive and may lead to serious visual impairment."

Seven affected individuals from four unrelated families, all homozygous for the same allele, ascertained as retinal dystrophy rather than as a glycosylation disorder. This is the basis for raising the frequency band and moving onset earlier.

PMID:28253385 SUPPORT Human Clinical

"electrodiagnostic testing revealed rod and cone dysfunction in the 5 patients tested"

Functional confirmation that both photoreceptor classes are involved, which is what makes the rod-cone dystrophy term the right binding.

PMID:31638560 SUPPORT Human Clinical

"showed clear signs of retinal dystrophy not recognized until our investigation"

A patient followed for years as congenital nystagmus and optic neuropathy whose retinal dystrophy was only found on dedicated imaging. Direct evidence that the occasional banding in earlier reports reflects ascertainment.

+ 1 more reference

Nyctalopia OCCASIONAL HP:0000662

Onset: CHILDHOOD

Show evidence (1 reference)

PMID:28253385 SUPPORT Human Clinical

"electrodiagnostic testing revealed rod and cone dysfunction in the 5 patients tested"

The rod dysfunction that night blindness reflects, measured in the same series that reports the symptom. The sentence listing the symptom itself is interrupted by bracketed measurement values, so the electrodiagnostic result is quoted instead. The band is OCCASIONAL because the symptom is described for the series as a whole without a count.

Myopia OCCASIONAL HP:0000545

Show evidence (1 reference)

PMID:31638560 SUPPORT Human Clinical

"with a myopic refractive error of -3.0 dioptre sphere"

A measured refractive error in a patient with SRD5A3-CDG and retinal dystrophy.

Head and Neck 1

Microcephaly OCCASIONAL HP:0000252

Show evidence (1 reference)

PMID:22304929 SUPPORT Human Clinical

"slight muscular hypotonia and microcephaly"

Microcephaly among the index patient's presenting features.

Integument 2

Ichthyosis OCCASIONAL HP:0008064

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"She was diagnosed with psoriasis as a child but recent dermatological evaluation is more consistent with diffuse ichthyosiform changes and palmoplantar keratoderma."

Documents the skin phenotype and the misdiagnosis that commonly precedes it.

Palmoplantar keratoderma OCCASIONAL HP:0000982

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"a more recent dermatological evaluation is more consistent with diffuse palmoplantar keratoderma"

Documents palmoplantar keratoderma in a reported sibling.

Musculoskeletal 2

Hypotonia FREQUENT HP:0001252

Show evidence (1 reference)

PMID:22304929 SUPPORT Human Clinical

"psychomotor retardation, pathological nystagmus, slight muscular hypotonia and microcephaly"

Hypotonia among the presenting features of the index patient.

Kyphosis OCCASIONAL HP:0002808

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"She is legally blind and has developed marked kyphosis and scoliosis as an adult."

Documents kyphosis as an acquired adult feature in a reported patient.

PMID:20700148 SUPPORT Human Clinical

"a novel syndrome consisting of mental retardation, coloboma, cataract and kyphosis"

Kyphosis is one of the four defining features of the independently described Kahrizi syndrome that this paper showed to be caused by SRD5A3, which is why it is curated here rather than treated as an incidental finding.

Nervous System 4

Intellectual disability VERY_FREQUENT HP:0001249

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"nearly every affected individual reported to date has cognitive delays, nystagmus/optic atrophy/optic hypoplasia, and ataxia"

"Nearly every affected individual" is the basis for the VERY_FREQUENT band.

Ataxia VERY_FREQUENT HP:0001251

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"and ataxia (some with documented vermis hypoplasia)"

Ataxia listed among the near-universal features in the same sentence as cognitive delay and the optic findings.

Cerebellar vermis hypoplasia FREQUENT HP:0001320

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"Head MRI found a hypoplastic inferior cerebellar vermis."

Documented imaging finding in a reported patient. The qualifier "some with documented vermis hypoplasia" elsewhere in the same paper is why this is banded below the ataxia itself.

PMID:41769439 REFUTE Human Clinical

"despite prominent neurological symptoms, brain MRIs at 20 months were entirely normal, showing no evidence of cerebellar hypoplasia or atrophy typically associated with this condition"

Two monozygotic twins with the recurrent nonsense allele, marked hypotonia, nystagmus and optic atrophy, and no cerebellar abnormality on imaging at 20 months. Cited as REFUTE against reading this finding as a constant, and because it adds a timing dimension the other reports do not: a normal early scan does not exclude the diagnosis.

Seizure OCCASIONAL HP:0001250

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"He started non-febrile seizures at 3 months old controlled with medication."

Documents early-onset epilepsy in a reported patient.

PMID:41769439 SUPPORT Human Clinical

"both twins exhibited generalized tonic-clonic seizures starting in early infancy - a feature less commonly reported in SRD5A3-CDG"

A second independent report of infantile-onset seizures, and the source that says explicitly that seizures are under-reported for this disorder. That is the reason the band is left at OCCASIONAL rather than raised: the authors are describing what the literature records, not a counted frequency.

Other 5

Optic nerve hypoplasia FREQUENT HP:0000609

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"In addition to optic nerve hypoplasia/atrophy, multiple individuals have been reported with colobomas of the optic nerve and iris"

Establishes optic nerve hypoplasia as a recurrent finding. The source groups hypoplasia with atrophy rather than counting them separately, so the band here is set one step below the combined figure.

Iris coloboma OCCASIONAL HP:0000612

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"multiple individuals have been reported with colobomas of the optic nerve and iris"

Establishes coloboma as recurrent but, in contrast to the optic nerve findings, not described as ubiquitous.

PMID:20700148 SUPPORT Human Clinical

"a novel syndrome consisting of mental retardation, coloboma, cataract and kyphosis"

Coloboma is one of the four defining features of Kahrizi syndrome, which this paper mapped to a homozygous SRD5A3 frameshift and showed to be allelic with SRD5A3-CDG.

Reduced protein C activity OCCASIONAL HP:0005543

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"Protein C was 50% (normal range 70-180%), protein S was normal at 83%, antithrombin III was 53%"

A measured protein C activity below the laboratory reference interval quoted alongside it in the same sentence, with protein S normal in the same patient.

Reduced antithrombin III activity OCCASIONAL HP:0001976

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"protein S was normal at 83%, antithrombin III was 53%"

A measured antithrombin III activity below its reference interval, with protein S normal in the same patient, so the deficiency is selective rather than a global loss of plasma protein.

Abnormal serum transferrin glycoform pattern Abnormal isoelectric focusing of serum transferrin HP:0003160

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"Carbohydrate deficient transferrin testing found: mono-oligo/di-oligo ratio of 0.44"

A measured abnormal transferrin glycoform ratio in a reported patient, with the normal cut-off quoted immediately after it in the source.

PMID:41732066 SUPPORT Human Clinical

"the current diagnosis of SRD5A3-CDG by screening assays remains challenging"

States that the existing screening assays are not adequate, which is the practical consequence of the partial hypoglycosylation.

🧬

Genetic Associations

1

SRD5A3 pathogenic variants (Causative)

Gene: SRD5A3 hgnc:25812 relationship_type: CAUSATIVE

Show evidence (6 references)

PMID:27480077 SUPPORT Human Clinical

"Sanger sequencing of SRD5A3 was performed and confirmed a novel homozygous nonsense mutation"

A documented homozygous null allele in affected siblings.

PMID:27480077 SUPPORT Human Clinical

"These two are of interest because they are among the first individuals reported with SRD5A3-CDG due to compound heterozygous mutations with one of the mutations being a missense mutation."

Establishes compound heterozygosity with a missense allele as the less common genotype, in the two patients with the mildest cognitive phenotype.

PMID:27480077 SUPPORT Human Clinical

"Patient 1 has more severe skin findings and marked short stature while her brother has an average stature and more severe structural eye abnormalities with retinal and iris colobomas."

Within-family divergence on an identical genotype, which is the evidence that genotype does not fix the phenotype here.

+ 3 more references

Variants (6)

SRD5A3 c.57G>A (p.Trp19X) Pathogenic

Gene: SRD5A3 hgnc:25812 NONSENSE

An exon 1 nonsense allele leaving no functional protein. It is the genotype behind the entry's central mechanistic observation: fibroblasts from this homozygous patient still contained normal dolichol while accumulating polyprenol, which is what forces the mechanism onto the polyprenol-to-dolichol ratio rather than onto dolichol depletion. It is also the most frequently reported allele in this disorder, and the reports that carry it are the ones that reframed the eye phenotype: seven individuals from four unrelated South Asian families, all homozygous for it, ascertained through eye clinics as early-onset retinal dystrophy, and separately a pair of monozygotic twins. Whether that reflects a founder haplotype is not established in the cited sources — the families are described as unrelated, and no haplotype analysis is reported — so it is recorded here as recurrent, not as a founder allele.

Functional effects

polyprenol reductase activity

Abolishes the enzyme. Despite that, dolichol levels are preserved, implying an alternative biosynthetic route.

Show evidence (3 references)

PMID:27480077 SUPPORT In Vitro

"evaluation of fibroblasts from a child with a homozygous stop codon in exon one at p.Trp19X in SRD5A3 found normal levels of dolichol but elevated polyprenol levels"

Names the allele and reports the lipid measurements made in cells carrying it.

PMID:28253385 SUPPORT Human Clinical

"identified the same homozygous SRD5A3 c.57G>A, p.(Trp19Ter) variant as the underlying cause of early-onset retinal dystrophy in each family"

The same allele in four unrelated families, which is the basis for calling it recurrent.

PMID:41769439 SUPPORT Human Clinical

"Whole-exome sequencing identified a homozygous nonsense mutation at exon 1 c.57G>A"

An independent report of the same allele, in the twins whose normal early brain imaging is cited against the vermis hypoplasia phenotype.

SRD5A3 c.603G>A (p.Trp201X) Pathogenic

Gene: SRD5A3 hgnc:25812 NONSENSE

A nonsense allele found homozygous in two affected siblings. Those siblings differ from each other in whether they have structural eye anomalies and in the severity of their skin involvement, which is this entry's evidence that genotype does not fix the phenotype.

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"Sanger sequencing of SRD5A3 was performed and confirmed a novel homozygous nonsense mutation (c.603G>A, p.Trp201X) in both siblings."

The allele and its homozygous state in both affected siblings.

SRD5A3 c.460T>C (p.Ser154Pro) Pathogenic

Gene: SRD5A3 hgnc:25812 MISSENSE

A missense allele modelled onto the predicted protein structure. It falls in a potential active site and is predicted to reduce catalytic efficiency rather than abolish it. Missense alleles are the minority in this disorder and are associated with the milder end of the cognitive spectrum, which is consistent with residual activity.

Functional effects

polyprenol reductase activity

Predicted to reduce catalytic efficiency. This is an in silico prediction from structural modelling, not a measured enzyme activity.

Show evidence (1 reference)

PMID:33403770 SUPPORT Computational

"Herein, we reported a novel SRD5A3 missense pathogenic variant c.460 T > C p.(Ser154Pro)."

The allele as reported. Its functional consequence is modelled rather than assayed, which is why the evidence is graded COMPUTATIONAL.

SRD5A3 c.203dupC (p.Phe69LeufsX2) Pathogenic

Gene: SRD5A3 hgnc:25812 FRAMESHIFT

A frameshift allele found homozygous in a consanguineous family whose condition had been described, and mapped, as a separate entity — Kahrizi syndrome, defined by intellectual disability with coloboma, cataract and kyphosis. Identifying this variant is what showed the two conditions to be the same gene, and is the reason Kahrizi syndrome is a synonym of this entry rather than a differential diagnosis.

Show evidence (2 references)

PMID:20700148 SUPPORT Human Clinical

"identified a homozygous frameshift mutation (c.203dupC; p.Phe69LeufsX2) in the gene for steroid 5α-reductase type 3 (SRD5A3) as the disease-causing change in this interval"

Names the allele and the mapped interval it was found in.

PMID:20700148 SUPPORT Human Clinical

"Our results show that Kahrizi syndrome and this CDG Ix subtype are allelic disorders"

The conclusion that merges the two named conditions onto one gene, stated by the authors who established it.

SRD5A3 c.436G>A (p.Glu146Lys) Pathogenic

Gene: SRD5A3 hgnc:25812 MISSENSE

A missense allele found homozygous in a child followed for years as congenital nystagmus with bilateral optic neuropathy, in whom dedicated retinal imaging then showed an unrecognised retinal dystrophy. The same patient had optic nerve head drusen, which the authors state had not previously been seen in this syndrome; that is a single observation and is not curated as a phenotype here.

Show evidence (1 reference)

PMID:31638560 SUPPORT Human Clinical

"a 12-year-old Czech child harbouring a novel homozygous variant, c.436G>A, p.(Glu146Lys) in SRD5A3"

The allele and the patient it was found in.

SRD5A3 c.509A>G (p.Tyr170Cys) Likely Pathogenic

Gene: SRD5A3 hgnc:25812 MISSENSE

A missense allele reported in a single patient who had the expected retinal dystrophy, ataxia and ichthyosiform skin change, together with a proximal limb-girdle pattern of weakness, brisk reflexes and extensor plantar responses. The authors propose those last findings as an expansion of the neurological phenotype. This entry records the allele but not the proposed new features as phenotypes — see the discussion on corticospinal involvement for why.

Show evidence (2 references)

PMID:41667393 SUPPORT Human Clinical

"uncovered a homozygous, likely pathogenic missense variant c.509A > G, p.(Tyr170Cys) in SRD5A3 gene"

The allele and the significance class the reporting authors assigned, which is why this variant is LIKELY_PATHOGENIC rather than PATHOGENIC.

PMID:41667393 SUPPORT Human Clinical

"plasma glycoprotein markers for N- and O-glycosylation showed an aberrant glycosylation profile"

Biochemical confirmation that the allele produces a glycosylation defect, which is what makes it more than a segregating rare variant.

💊

Medical Actions

3

Atorvastatin (experimental, not established therapy)

Action: Pharmacotherapy NCIT:C15986

Agent: atorvastatin CHEBI:39548

The only candidate disease-modifying therapy proposed for this disorder, and it is at the preclinical stage. An HMG-CoA reductase inhibitor is a counter-intuitive choice here - it restricts the isoprenoid supply upstream of polyprenol - but the logic follows from this entry's mechanism: if the pathogenic quantity is the polyprenol-to-dolichol ratio rather than the dolichol level, then reducing polyprenol production should help even though it reduces flux into the pathway overall. In patient fibroblasts it did restore the ratio, and in the worm model it rescued the phenotype. Recorded here so the reasoning is visible, explicitly NOT as a treatment in clinical use: no patient has been treated, and no trial is reported.

Mechanism Target:

Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio — The drug acts on the node this entry treats as the operative lesion, and the reported readout is that node's own measurement: the polyprenol-to-dolichol ratio in patient fibroblasts.

Show evidence (2 references)

PMID:41648237 SUPPORT In Vitro

"restored polyprenol-to-dolichol ratios in patient fibroblasts"

The correction of the ratio in patient cells, which is what makes the drug act on this node. The strongest independent support for the ratio model in this entry, because it shows the ratio is pharmacologically movable.

PMID:41648237 SUPPORT Model Organism

"Atorvastatin rescued disease-relevant phenotypes in the worm model"

The organismal half of the same result: moving the ratio changes phenotype. Split from the fibroblast clause because the two halves of that sentence report different kinds of study.

Show evidence (3 references)

PMID:41648237 SUPPORT Model Organism

"Atorvastatin rescued disease-relevant phenotypes in the worm model"

The whole-organism half of the preclinical result. No human administration is reported, which is why this entry names itself experimental.

PMID:41648237 SUPPORT In Vitro

"restored polyprenol-to-dolichol ratios in patient fibroblasts"

The cell-culture half of the same result. Kept as a separate item because evidence_source classifies the study, and one sentence here reports two.

PMID:41648237 SUPPORT Human Clinical

"Approximately 60 cases have been reported, with treatment limited to symptomatic management."

States the current standard of care against which this candidate is proposed: symptomatic management only.

Ophthalmological surveillance and management

Action: Supportive Care NCIT:C15747

No disease-modifying therapy exists. Because cataract, retinitis pigmentosa and glaucoma develop over time in patients diagnosed as children, lifelong ophthalmological monitoring is recommended, with intervention for the treatable components.

Target Phenotypes: Cataract HP:0000518 Retinal dystrophy HP:0000556

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"All individuals diagnosed with SRD5A3-CDG as children will need close monitoring for the development of cataracts, retinitis pigmentosa, and kyphosis as they age."

The explicit surveillance recommendation this entry implements.

Developmental and neurological supportive care

Action: Supportive Care NCIT:C15747

Management of the developmental delay, ataxia and epilepsy: anticonvulsants where seizures occur, feeding support where oral intake is inadequate, and developmental therapies. Physical, occupational and speech therapy and orthopaedic management of the kyphosis are all reasonable and are named in review prose, but no cited source documents them in an SRD5A3-CDG patient, so they are not modelled as separate treatment entries here. That is the position for the CDGs generally rather than a gap peculiar to this disorder: symptomatic management is what most of them have, and the exceptions are the few with a dietary sugar therapy.

Target Phenotypes: Seizure HP:0001250 Ataxia HP:0001251

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"He had difficulty taking sufficient calories orally and a G-tube was placed providing all nutrition."

Documents the supportive interventions actually used in a reported patient.

PMID:39236565 SUPPORT Other

"Mostly, we are only able to manage the disease symptoms rather than to address the underlying cause."

Places this entry's supportive-only treatment section in its context across the CDGs. Graded OTHER because the source is a treatment overview rather than a study reporting data.

🔬

Biochemical Markers

1

Elevated polyprenol-to-dolichol ratio (INCREASED)

Context: Measured by lipid quantification in cultured patient fibroblasts. NCIT has no term for either lipid - searching it returns the reductase protein and the SRD5A3 allele, not the analytes - so this biomarker is deliberately left with a free-text preferred_term rather than bound to an unrelated code.

Pathograph Readouts

Readout Of Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio Positive Diagnostic

The ratio is the direct measurement of the node it reports on, in the tissue the measurement was made in.

Show evidence (1 reference)

PMID:22304929 SUPPORT In Vitro

"demonstrated a high polyprenol/dolichol ratio with normal amounts of dolichol"

The raised ratio in patient cells, which is the node's own defining quantity.

Pharmacodynamic Marker Of SRD5A3 Polyprenol Reductase Deficiency Positive Pharmacodynamic

Correction of the ratio in patient fibroblasts is the readout the one candidate therapy was assessed against, so it doubles as a pharmacodynamic marker for any future attempt to treat the enzymatic lesion.

Show evidence (1 reference)

PMID:41648237 SUPPORT In Vitro

"restored polyprenol-to-dolichol ratios in patient fibroblasts"

The ratio used as the drug-response endpoint in patient cells.

Show evidence (2 references)

PMID:22304929 SUPPORT In Vitro

"Quantification of dolichol and unreduced polyprenol in the patient's fibroblasts demonstrated a high polyprenol/dolichol ratio with normal amounts of dolichol"

The measurement itself, including the normal dolichol level that makes the ratio rather than the level the informative quantity.

PMID:20637498 SUPPORT In Vitro

"The presence of residual dolichol in cells depleted for this enzyme suggests the existence of an unexpected alternative pathway for dolichol de novo biosynthesis."

Independent confirmation from the gene-discovery paper that dolichol is not depleted, which is what a laboratory interpreting a dolichol level alone would misread.

🔬

Diagnosis

4

Transferrin screening followed by SRD5A3 sequencing

Suspicion is raised by the combination of developmental delay, nystagmus with an optic nerve abnormality and ataxia. Carbohydrate-deficient transferrin testing showing a type 1 pattern places the disorder among the CDG type I group, and sequencing of SRD5A3 confirms it. The screening step is imperfect, because hypoglycosylation is partial even in patients with no functional enzyme, so a normal or near-normal transferrin profile does not exclude the diagnosis.

carbohydrate-deficient transferrin measurement NCIT:C101016

Show evidence (2 references)

PMID:27480077 SUPPORT Human Clinical

"Despite the normal levels of dolichol, there was hypoglycosylation with only 70% of transferrin in serum correctly glycosylated."

Quantifies how mild the screening abnormality can be in a patient with a complete enzyme null, which is why the screen is not a rule-out.

PMID:41732066 SUPPORT Human Clinical

"Some of these alterations could be further pursued to develop glycopeptide-based biomarkers as the current diagnosis of SRD5A3-CDG by screening assays remains challenging."

States the diagnostic gap and the proposed direction for closing it.

Dysmorphology-assisted recognition

Recurrent facial features have been catalogued and used to build an automated 2D facial-image classifier for SRD5A3-CDG. The reported accuracy is from the developers' own validation on published and in-house cases; it has not been independently replicated, so it is recorded here as a proposed adjunct rather than an established diagnostic test.

Show evidence (2 references)

PMID:33403770 SUPPORT Computational

"Based on facial digital 2D images, we successfully designed and validated a SRD5A3-CDG computer based dysmorphic facial analysis, which achieved 92.5% accuracy."

The reported performance of the classifier, from the group that built it.

PMID:33403770 SUPPORT Human Clinical

"some recurrent dysmorphic features such as arched eyebrows, wide eyes, shallow nasal bridge, short nose, and large mouth"

The dysmorphic features the classifier and clinical recognition both rest on.

Retinal electrophysiology and imaging

The route by which this diagnosis is actually reached in a substantial share of patients, and the one the transferrin screen above does not cover. Fundus autofluorescence and optical coherence tomography detect the retinal dystrophy, and electroretinography characterises it as combined rod and cone dysfunction. Two things make this more than a supporting investigation. It has found retinal dystrophy in a patient followed for years under a different ophthalmic diagnosis, so a normal-looking fundus is not a rule-out; and in one series every patient came to attention this way rather than through metabolic medicine. NCIT has no electroretinography term reachable from Clinical Intervention or Procedure, so the binding names the imaging half and the preferred term carries both.

fundus autofluorescence imaging with electroretinography NCIT:C162465

Show evidence (2 references)

PMID:28253385 SUPPORT Human Clinical

"Fundus autofluorescence imaging and optical coherence tomographic scans were abnormal in all patients, and electrodiagnostic testing revealed rod and cone dysfunction in the 5 patients tested."

The yield of this workup in the series where it was applied systematically.

PMID:31638560 SUPPORT Human Clinical

"Examination by spectral domain optical coherence tomography and fundus autofluorescence imaging showed clear signs of retinal dystrophy not recognized until our investigation"

A retinal dystrophy missed for years and found only when this workup was done, which is why it is curated as a diagnostic step rather than as characterisation.

Intragenic copy-number analysis when sequencing is uninformative

Sequencing does not always find the second allele. In one reported sibship the second variant was an intragenic tandem duplication of exons 2 to 4, invisible to the standard exome pipeline and found only when copy-number analysis was applied to the exome data. The patients were clinically typical, so nothing about the presentation signalled that the first result was incomplete. This is recorded as a diagnostic step so that a single heterozygous pathogenic variant in a clinically compatible patient is not read as excluding the diagnosis. NCIT has no term for this analysis under Clinical Intervention or Procedure, so it is left with a free-text preferred term.

exome-based copy number analysis

Show evidence (1 reference)

PMID:35339718 SUPPORT Human Clinical

"Only when applying exome-based copy number analysis, we identified as a second compound heterozygous variant a previously not reported tandem duplication of exons 2-4 in SRD5A3."

The allele and the method that found it, in patients whose sequencing had been uninformative.

📊

Prevalence

1

Global published literature

Cases In Literature Ultra Rare

A diagnosed-case count as of 2021, not a population prevalence estimate. No population-based rate has been published for this disorder.

Show evidence (3 references)

PMID:33403770 SUPPORT Human Clinical

"To date, 43 affected individuals have been reported."

A directly stated cumulative published patient count as of 2021.

PMID:41648237 SUPPORT Human Clinical

"Approximately 60 cases have been reported, with treatment limited to symptomatic management."

The current count, four years later. Roughly 60 against 43 gives some sense of the reporting rate for this disorder: about four new cases a year.

PMID:35339718 SUPPORT Human Clinical

"So far, only 23 distinct mutations were described."

The size of the allelic spectrum as of 2022, which bounds this entry's variants list as a sample rather than a catalogue.

🔀

Differential Diagnoses

1

Conditions with similar clinical presentations that must be differentiated from SRD5A3-Congenital Disorder of Glycosylation:

DOLK-congenital disorder of glycosylation

Overlapping Features The disorder of the immediately adjacent step, in which dolichol is converted to dolichol phosphate. It shares the ichthyosiform skin change, seizures and hypotonia but characteristically causes cardiomyopathy and does not produce the prominent eye involvement that defines SRD5A3-CDG. The contrast is mechanistically informative: it is what makes the ocular phenotype look like a consequence of polyprenol accumulation rather than of downstream hypoglycosylation alone.

Show evidence (1 reference)

PMID:27480077 SUPPORT Human Clinical

"the conversion of dolichol to dolichol phosphate by DOLK leads to cardiomyopathy, seizures, hypotonia, and ichthyosiform skin changes, but significant eye abnormalities have not been reported"

The direct comparison between the two adjacent pathway defects, including the absence of eye involvement in the downstream one.

🐁

Animal Models

1

C. elegans SRD5A3 W19X model

The first disease model for SRD5A3-CDG, and the reason a therapeutic candidate exists at all. Worms homozygous for the W19X nonsense allele - the same allele carried by the patient whose fibroblast lipid measurements underpin this entry's mechanism - reproduce developmental delay, neurological dysfunction and mevalonate pathway dysregulation. It was built specifically to be screenable, which a mouse model would not be even if one existed. The choice of allele matters more than "commonly observed in patients" conveys. W19X is the most frequently reported allele in this disorder: it is the homozygous genotype of all seven patients in the four-family retinal-dystrophy series and of the reported twin pair, as well as of the fibroblast donor. So the model is built on the allele that the largest share of published patients actually carry, rather than on a convenient null.

Species

Caenorhabditis elegans

Genotype

srd-5a3 W19X homozygous (patient nonsense allele knock-in)

Publication

PMID:41648237

Pathograph links

SRD5A3 Polyprenol Reductase Deficiency recapitulates fidelity: moderate Carries a patient allele and reproduces disease-relevant phenotypes at the organismal level, including a mevalonate-pathway readout that connects the lesion to the isoprenoid supply upstream of polyprenol.

Limitations: An invertebrate. It cannot model the features that define the human disease clinically - the ocular malformations, the cerebellar vermis hypoplasia, the skin phenotype - so its value is as a screening system and a mechanism probe, not as a phenocopy.

Readouts

Motility in a high-throughput screen

Direction: DECREASED

Interpretation: The screenable phenotype the drug repurposing screen was run against.

Show evidence (1 reference)

PMID:41648237 SUPPORT Model Organism

"Using this model, we conducted a high-throughput motility-based drug repurposing screen and identified atorvastatin, an FDA-approved HMG-CoA reductase inhibitor, as a repurposing candidate."

Names the readout and what it was used for.

Show evidence (1 reference)

PMID:41648237 SUPPORT Model Organism

"we developed the first C. elegans model of SRD5A3-CDG, harboring the homozygous W19X nonsense mutation commonly observed in patients. This model recapitulates disease-relevant phenotypes, including developmental delays, neurological dysfunction, and mevalonate pathway dysregulation."

The model, its genotype, and the phenotypes it reproduces.

{ }

Source YAML

click to show

name: SRD5A3-Congenital Disorder of Glycosylation
creation_date: "2026-09-01T14:07:00Z"
description: >-
  SRD5A3-CDG (formerly CDG-Iq) is an autosomal recessive congenital disorder of
  glycosylation caused by biallelic loss of SRD5A3, which encodes polyprenol
  reductase. The enzyme reduces the alpha-isoprene unit of polyprenol to form
  dolichol, the lipid carrier on which the oligosaccharide precursor for N-linked
  glycosylation is assembled and which is also required for O-mannosylation,
  C-mannosylation and GPI anchor synthesis. Because the block sits at the very
  start of the pathway, one enzymatic lesion produces hypoglycosylation of a wide
  range of secreted and membrane proteins. The clinical picture is dominated by a
  strikingly consistent neuro-ophthalmological core - cognitive delay, nystagmus
  with optic nerve hypoplasia or atrophy, and ataxia, often with cerebellar vermis
  hypoplasia - on which more variable ocular structural defects (iris and optic
  nerve coloboma), ichthyosiform skin change with palmoplantar keratoderma,
  coagulation factor deficiencies and, in adults, cataract, retinitis pigmentosa
  and kyphosis are superimposed. The mechanism is not a simple loss of dolichol:
  patient fibroblasts carrying a null allele retain normal dolichol levels while
  accumulating unreduced polyprenol, so the operative lesion is most likely the
  raised polyprenol-to-dolichol ratio rather than dolichol depletion.
synonyms:
- SRD5A3-CDG
- congenital disorder of glycosylation type Iq
- CDG1Q
- congenital disorder of glycosylation due to steroid 5alpha-reductase type 3 deficiency
- polyprenol reductase deficiency
- Kahrizi syndrome
category: Mendelian
disease_term:
  preferred_term: SRD5A3-congenital disorder of glycosylation
  term:
    id: MONDO:0012885
    label: SRD5A3-congenital disorder of glycosylation
mappings:
  mondo_mappings:
  - term:
      id: MONDO:0012885
      label: SRD5A3-congenital disorder of glycosylation
    mapping_predicate: skos:exactMatch
    mapping_source: MONDO
parents:
- Congenital disorder of glycosylation type I
- Inborn error of metabolism
inheritance:
- name: Autosomal recessive inheritance
  inheritance_term:
    preferred_term: Autosomal recessive inheritance
    term:
      id: HP:0000007
      label: Autosomal recessive inheritance
  description: >-
    Affected individuals carry biallelic SRD5A3 variants. Most reported patients
    are homozygous for nonsense or frameshift alleles and come from consanguineous
    families; compound heterozygosity including a missense allele is less common
    and was among the first observed in adults with a milder cognitive phenotype.
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The majority of individuals with SRD5A3-CDG reported in the literature have
      homozygote or compound heterozygote nonsense or frame shift mutations, most
      with documented consanguinity
    explanation: >-
      States the biallelic requirement and the predominance of homozygous
      loss-of-function alleles arising through consanguinity.
prevalence:
- population: Global published literature
  measure_type: CASES_IN_LITERATURE
  prevalence_class: ULTRA_RARE
  notes: >-
    A diagnosed-case count as of 2021, not a population prevalence estimate. No
    population-based rate has been published for this disorder.
  evidence:
  - reference: PMID:33403770
    reference_title: "SRD5A3-CDG: 3D structure modeling, clinical spectrum, and computer-based dysmorphic facial recognition."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      To date, 43 affected individuals have been reported.
    explanation: >-
      A directly stated cumulative published patient count as of 2021.
  - reference: PMID:41648237
    reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Approximately 60 cases have been reported, with treatment limited to symptomatic
      management.
    explanation: >-
      The current count, four years later. Roughly 60 against 43 gives some sense of the
      reporting rate for this disorder: about four new cases a year.
  - reference: PMID:35339718
    reference_title: "SRD5A3-CDG: Twins with an intragenic tandem duplication."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      So far, only 23 distinct mutations were described.
    explanation: >-
      The size of the allelic spectrum as of 2022, which bounds this entry's variants list
      as a sample rather than a catalogue.
mechanistic_hypotheses:
- hypothesis_group_id: polyprenol_competition_model
  hypothesis_label: Raised Polyprenol-to-Dolichol Ratio Model
  status: CANONICAL
  description: >-
    The operative lesion is not the absence of dolichol but the accumulation of its
    unreduced precursor. Fibroblasts from a patient homozygous for an exon 1 stop
    codon, with no functional polyprenol reductase, contained normal amounts of
    dolichol together with elevated polyprenol, yet still hypoglycosylated
    transferrin. On this model the excess polyprenol competes with dolichol at the
    initiation of N-glycan assembly without being able to support glycosylation, so
    the pathogenic quantity is the ratio rather than the dolichol level. It also
    accounts for why complete dolichol loss is not seen: an alternative dolichol
    biosynthetic route persists, and its absence would be expected to be lethal.
  evidence:
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Quantification of dolichol and unreduced polyprenol in the patient's
      fibroblasts demonstrated a high polyprenol/dolichol ratio with normal amounts
      of dolichol, indicating that high polyprenol levels might compete with dolichol
      for the initiation of N-glycan assembly but without supporting normal
      glycosylation
    explanation: >-
      The measurement and the interpretation this hypothesis is built on, stated by
      the authors who made it.
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      hypothesized that one explanation may be not the absence of dolichol (which
      would be lethal) but the increased ratio of polyprenol to dolichol leading to
      the phenotype associated with SRD5A3-CDG
    explanation: >-
      An independent restatement of the model, including the argument that dolichol
      absence would not be survivable.
  notes: >-
    Recorded as CANONICAL because it is the only mechanism proposed in this
    literature and is consistent with the fibroblast measurements, not because it has
    been tested against an alternative. The competition step itself is inferred from
    the lipid ratio; no assay of polyprenol-linked glycan initiation is cited.
pathophysiology:
- name: SRD5A3 Polyprenol Reductase Deficiency
  biological_scale: MOLECULAR
  description: >-
    Biallelic SRD5A3 variants abolish or reduce polyprenol reductase, the enzyme
    that reduces the alpha-isoprene unit of polyprenol to yield dolichol. Reported
    alleles are predominantly nonsense and frameshift changes; a missense allele
    modelled onto the predicted protein structure falls in a potential active site
    and is predicted to reduce catalytic efficiency rather than abolish it, which
    fits the milder phenotype seen in compound heterozygotes.

    What that enzymatic step is has been revised since SRD5A3 was identified. Finding the
    gene, together with its yeast counterpart, established the view that dolichol is made
    from polyprenol in one step; the later discovery of DHRSX in CDG patients replaced
    that with a three-step detour pathway, now shown to be conserved in yeast. The
    revision matters for this entry in a specific way: the yeast DHRSX counterpart cannot
    be substituted for by SRD5A3, so the two are separate, non-redundant steps rather
    than alternative routes to the same product. Whatever preserves dolichol levels in
    SRD5A3-null patient fibroblasts, it is not DHRSX standing in for the missing enzyme.
  genes:
  - preferred_term: SRD5A3
    term:
      id: hgnc:25812
      label: SRD5A3
  molecular_functions:
  - preferred_term: polyprenol reductase activity
    term:
      id: GO:0102389
      label: polyprenol reductase activity
    modifier: DECREASED
  downstream:
  - target: Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio
    causal_link_type: DIRECT
    description: >-
      Loss of the reductase leaves its substrate unconsumed, which is what raises the
      polyprenol-to-dolichol ratio.
    evidence:
    - reference: PMID:22304929
      reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
      supports: SUPPORT
      evidence_source: IN_VITRO
      snippet: >-
        Quantification of dolichol and unreduced polyprenol in the patient's
        fibroblasts demonstrated a high polyprenol/dolichol ratio with normal amounts
        of dolichol
      explanation: >-
        The direct measurement of substrate accumulation in patient cells.
  evidence:
  - reference: PMID:20637498
    reference_title: SRD5A3 is required for converting polyprenol to dolichol and is mutated in a congenital glycosylation disorder.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      We describe a new type of CDG caused by mutations in the steroid 5alpha-reductase
      type 3 (SRD5A3) gene.
    explanation: >-
      The report establishing SRD5A3 as the causal gene for this CDG.
  - reference: PMID:20637498
    reference_title: SRD5A3 is required for converting polyprenol to dolichol and is mutated in a congenital glycosylation disorder.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      We found that SRD5A3 is necessary for the reduction of the alpha-isoprene unit
      of polyprenols to form dolichols, required for synthesis of dolichol-linked
      monosaccharides, and the oligosaccharide precursor used for N-glycosylation.
    explanation: >-
      Identifies the exact enzymatic step lost, and why losing it reaches
      N-glycosylation.
  - reference: PMID:33403770
    reference_title: "SRD5A3-CDG: 3D structure modeling, clinical spectrum, and computer-based dysmorphic facial recognition."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      predicted that the p.(Ser154Pro) variant is located in a potential active site
      and is capable of reducing its catalytic efficiency
    explanation: >-
      Structural modelling of a missense allele. This is an in silico prediction, not
      a measured enzyme activity, and is graded accordingly.
  - reference: PMID:42201967
    reference_title: The revised three-step detour pathway in dolichol biosynthesis is evolutionarily conserved in budding yeast.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      a recent discovery of DHRSX in CDG patients revised this view and led to the
      proposal of a three-step detour pathway for dolichol biosynthesis
    explanation: >-
      Records that the one-step model this entry's enzymatic description was originally
      written against has been superseded, which is why the node describes the pathway
      context rather than only the reaction.
  - reference: PMID:42201967
    reference_title: The revised three-step detour pathway in dolichol biosynthesis is evolutionarily conserved in budding yeast.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      All these phenotypes were rescued by expression of DHRSX, but not by DFG10 or
      SRD5A3.
    explanation: >-
      A rescue experiment showing the two enzymes are not interchangeable. This is the
      basis for the node's statement that DHRSX is not the route that preserves dolichol
      in SRD5A3-null cells, and it constrains the entry's canonical hypothesis, which
      leaves that route unidentified.
- name: Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio
  biological_scale: MOLECULAR
  description: >-
    Unreduced polyprenol accumulates while dolichol levels remain near normal,
    because an alternative dolichol biosynthetic route persists. The resulting shift
    in the ratio, rather than a shortage of dolichol, is the proposed proximate cause
    of hypoglycosylation: excess polyprenol is thought to compete with dolichol at
    the initiation of glycan assembly without being able to carry the glycan.
  downstream:
  - target: Impaired Dolichol-Linked Oligosaccharide Assembly
    causal_link_type: DIRECT
    hypothesis_groups:
    - polyprenol_competition_model
    description: >-
      Competition by the accumulated precursor at the initiation step is the proposed
      route from the lipid imbalance to defective oligosaccharide assembly.
    evidence:
    - reference: PMID:22304929
      reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
      supports: SUPPORT
      evidence_source: IN_VITRO
      snippet: >-
        high polyprenol levels might compete with dolichol for the initiation of
        N-glycan assembly but without supporting normal glycosylation
      explanation: >-
        The proposed competition step. The authors' hedge is preserved: this is a
        mechanism inferred from the lipid ratio, not a measured competition.
      directness: INDIRECT
  evidence:
  - reference: PMID:20637498
    reference_title: SRD5A3 is required for converting polyprenol to dolichol and is mutated in a congenital glycosylation disorder.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      The presence of residual dolichol in cells depleted for this enzyme suggests the
      existence of an unexpected alternative pathway for dolichol de novo biosynthesis.
    explanation: >-
      Explains why dolichol is not simply absent, which is what forces the mechanism
      onto the ratio rather than onto depletion.
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      SRD5A3 is not the sole producer of dolichol since evaluation of fibroblasts from
      a child with a homozygous stop codon in exon one at p.Trp19X in SRD5A3 found
      normal levels of dolichol but elevated polyprenol levels
    explanation: >-
      Independent statement that a complete null allele still leaves normal dolichol,
      with elevated polyprenol.
- name: Impaired Dolichol-Linked Oligosaccharide Assembly
  biological_scale: MOLECULAR
  description: >-
    Assembly of the lipid-linked oligosaccharide precursor on dolichol is
    compromised. Because dolichol also serves O-mannosylation, C-mannosylation and
    GPI anchor synthesis, the defect is not confined to the N-linked pathway, which
    is part of why the phenotype spans so many organ systems.
  biological_processes:
  - preferred_term: dolichol-linked oligosaccharide biosynthetic process
    term:
      id: GO:0006488
      label: dolichol-linked oligosaccharide biosynthetic process
    modifier: DECREASED
  downstream:
  - target: Systemic Protein Hypoglycosylation
    causal_link_type: DIRECT
    description: >-
      A defective precursor is transferred to fewer nascent glycoproteins, which is
      what hypoglycosylation consists of.
    evidence:
    - reference: PMID:41732066
      reference_title: Extensive Hypoglycosylation of Serum N-Glycoproteins in SRD5A3 Deficiency.
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Extensive hypoglycosylation of serum proteins was observed in patients, with
        245 of 291 altered glycopeptides decreased in SRD5A3-CDG.
      explanation: >-
        Quantifies the hypoglycosylation directly at the glycopeptide level in patient
        serum.
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This is a major pathway for dolichol biosynthesis for N-glycosylation,
      O-mannosylation, C-mannosylation, and GPI anchor synthesis.
    explanation: >-
      Establishes that the affected carrier serves four glycosylation pathways, not
      only the N-linked one.
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      an enzyme catalyzing the final step of the biosynthesis of dolichol, which is
      required for the assembly of the glycans needed for N-glycosylation
    explanation: >-
      Places the enzymatic block immediately upstream of glycan assembly.
- name: Systemic Protein Hypoglycosylation
  biological_scale: ORGANISM
  description: >-
    Secreted and membrane glycoproteins carry fewer or truncated N-glycans. Serum
    glycoproteomics in patients shows the change is broad rather than selective, and
    the affected proteins include several whose deficiency has recognised clinical
    consequences - haptoglobin, plasma serine protease inhibitor, alpha-1-B
    glycoprotein, alpha-2-macroglobulin and ceruloplasmin. Hypoglycosylation is
    partial rather than complete: a patient with no functional enzyme still
    glycosylated most of their transferrin correctly.
  biological_processes:
  - preferred_term: protein N-linked glycosylation
    term:
      id: GO:0006487
      label: protein N-linked glycosylation
    modifier: DECREASED
  downstream:
  - target: Cerebellar and Optic Pathway Vulnerability
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      The cerebellum and optic tract are disproportionately affected in this CDG. The
      intervening steps between generalised hypoglycosylation and that selectivity are
      not established, which is why the link is marked as having unknown
      intermediates.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        The cerebellum and optic tract seem especially vulnerable to deficiency of
        SRD5A3.
      explanation: >-
        States the tissue selectivity that this edge records. The authors hedge
        ("seem"), and offer no mechanism for it.
      directness: INDIRECT
  - target: Abnormal serum transferrin glycoform pattern
    causal_link_type: DIRECT
    description: >-
      The transferrin glycoform profile is the direct laboratory readout of the
      hypoglycosylation, which is what makes it the screening test for this group of
      disorders.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Despite the normal levels of dolichol, there was hypoglycosylation with only 70%
        of transferrin in serum correctly glycosylated.
      explanation: >-
        Ties the measured transferrin abnormality directly to the hypoglycosylation.
  - target: Progressive Ocular and Skeletal Degeneration
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      The adult-onset features arise on the same biochemical background, but nothing in
      the cited literature explains why they are progressive rather than congenital, or
      why the retina and spine in particular. Drawn with unknown intermediates rather than
      left as a second root, because a second root would imply an independent cause.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        As highlighted by patients 4 and 5, individuals with SRD5A3-CDG have issues that
        can develop over time.
      explanation: >-
        Establishes that the late features belong to the same disease course rather than
        to a separate process, which is what this edge asserts. The mechanism of the
        progression is not addressed.
      directness: INDIRECT
  - target: Ichthyosis
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Ichthyosiform skin change occurs in SRD5A3-CDG and also in DOLK-CDG, the disorder of
      the immediately adjacent pathway step, which suggests it belongs to the shared
      glycosylation defect rather than to polyprenol accumulation specifically. The route
      from hypoglycosylation to the epidermal barrier is not established here.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        the conversion of dolichol to dolichol phosphate by DOLK leads to cardiomyopathy,
        seizures, hypotonia, and ichthyosiform skin changes
      explanation: >-
        The shared skin phenotype between two adjacent pathway defects, which is the basis
        for attributing it to the common downstream hypoglycosylation.
      directness: INDIRECT
  - target: Palmoplantar keratoderma
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Part of the same dermatological phenotype as the ichthyosiform change, and attributed
      the same way.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        recent dermatological evaluation is more consistent with diffuse ichthyosiform
        changes and palmoplantar keratoderma
      explanation: >-
        The two skin findings are reported together in the same patient, as one
        dermatological phenotype.
      directness: INDIRECT
  - target: Reduced protein C activity
    causal_link_type: INDIRECT_KNOWN_INTERMEDIATES
    description: >-
      Coagulation regulatory proteins are N-glycosylated, and reduced activity of
      protein C and antithrombin III is documented in patients.
    evidence:
    - reference: PMID:41732066
      reference_title: Extensive Hypoglycosylation of Serum N-Glycoproteins in SRD5A3 Deficiency.
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Some of these proteins have previously been reported to be associated with
        liver dysfunction, anemia, and coagulopathy, which could underlie similar
        clinical features observed in SRD5A3-CDG patients.
      explanation: >-
        Connects the measured serum hypoglycosylation to the coagulopathy. The authors
        write "could underlie", so this is a proposed rather than demonstrated route.
      directness: INDIRECT
  - target: Reduced antithrombin III activity
    causal_link_type: INDIRECT_KNOWN_INTERMEDIATES
    description: >-
      The same route as for protein C; both anticoagulant proteins were reduced in the
      reported siblings.
    evidence:
    - reference: PMID:41732066
      reference_title: Extensive Hypoglycosylation of Serum N-Glycoproteins in SRD5A3 Deficiency.
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Altered glycopeptides included those derived from haptoglobin, plasma serine
        protease inhibitor, alpha-1-B glycoprotein, alpha-2-macroglobulin, and
        ceruloplasmin.
      explanation: >-
        Names the affected serum glycoproteins, including a protease inhibitor class,
        which is the basis for attributing the anticoagulant deficiencies to
        hypoglycosylation.
      directness: INDIRECT
  evidence:
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      about 70% of transferrin (Tf) was correctly glycosylated
    explanation: >-
      Establishes that hypoglycosylation is partial even with a complete enzyme null,
      which constrains any model of the mechanism.
- name: Cerebellar and Optic Pathway Vulnerability
  biological_scale: TISSUE
  description: >-
    The cerebellum and the optic pathway are the tissues most consistently affected.
    Nearly every reported patient has cognitive delay, an optic nerve abnormality with
    nystagmus, and ataxia, frequently with cerebellar vermis hypoplasia. The optic
    involvement is not purely a developmental malformation: it also progresses,
    implicating long-term maintenance of the nerve as well as its formation. One
    proposed explanation for the selectivity is particular sensitivity of the optic
    nerve to polyprenol accumulation, which would distinguish this CDG from
    N-glycosylation defects further downstream.
  downstream:
  - target: Intellectual disability
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Cognitive delay is among the near-universal features.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        nearly every affected individual reported to date has cognitive delays
      explanation: >-
        Establishes cognitive delay as effectively constant across reported patients.
  - target: Ataxia
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Cerebellar dysfunction manifests as ataxia, often with documented vermis
      hypoplasia.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        and ataxia (some with documented vermis hypoplasia)
      explanation: >-
        Ataxia among the near-universal features, with the structural correlate.
  - target: Optic atrophy
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Optic nerve degeneration, distinct from the developmental hypoplasia and
      progressive over time.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        This does not appear to be a solely structural defect due to damage to the
        developing optic tract, but also impacts on long term optic nerve maintenance.
      explanation: >-
        Distinguishes the degenerative component from the malformative one, which is
        why atrophy and hypoplasia are curated as separate phenotypes.
  - target: Optic nerve hypoplasia
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      The developmental, as opposed to degenerative, form of the optic nerve involvement.
      The source groups the two and describes the combination as ubiquitous.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        In addition to optic nerve hypoplasia/atrophy, multiple individuals have been
        reported with colobomas of the optic nerve and iris
      explanation: >-
        Establishes optic nerve hypoplasia as part of the optic pathway phenotype.
  - target: Iris coloboma
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      A structural malformation of the developing eye, reported alongside optic nerve
      coloboma. It is a closure defect rather than a degenerative change, so it belongs to
      the developmental half of the ocular phenotype.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        multiple individuals have been reported with colobomas of the optic nerve and iris
      explanation: >-
        The coloboma phenotype, recurrent but not universal.
  - target: Cerebellar vermis hypoplasia
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      The structural correlate of the ataxia, present in a proportion of patients.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        and ataxia (some with documented vermis hypoplasia)
      explanation: >-
        The vermis hypoplasia is reported as accompanying the ataxia in some patients, not
        all, which is why it is banded below the ataxia itself.
  - target: Hypotonia
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Muscular hypotonia accompanies the cerebellar and wider neurological involvement.
    evidence:
    - reference: PMID:22304929
      reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        The clinical features were psychomotor retardation, pathological nystagmus, slight
        muscular hypotonia and microcephaly.
      explanation: >-
        Hypotonia grouped with the other neurological features in the index patient.
      directness: INDIRECT
  - target: Microcephaly
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Reduced head growth in a subset of patients, part of the neurological phenotype.
    evidence:
    - reference: PMID:22304929
      reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        pathological nystagmus, slight muscular hypotonia and microcephaly
      explanation: >-
        Microcephaly among the presenting neurological features.
      directness: INDIRECT
  - target: Seizure
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Epilepsy occurs in a subset, in some from early infancy. As with the ichthyosis, the
      same feature occurs in DOLK-CDG one step downstream, so it is more likely a
      consequence of hypoglycosylation than of polyprenol accumulation.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        He started non-febrile seizures at 3 months old controlled with medication.
      explanation: >-
        Documented early-onset epilepsy in a reported patient.
      directness: INDIRECT
  - target: Nystagmus
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Nystagmus accompanies the optic pathway involvement and is often the presenting
      sign in infancy.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        nearly every affected individual reported to date has cognitive delays,
        nystagmus/optic atrophy/optic hypoplasia, and ataxia
      explanation: >-
        Groups nystagmus with the optic findings among the near-universal features.
  evidence:
  - reference: PMID:20637498
    reference_title: SRD5A3 is required for converting polyprenol to dolichol and is mutated in a congenital glycosylation disorder.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Patients have mental retardation and ophthalmologic and cerebellar defects.
    explanation: >-
      The original description of the neuro-ophthalmological core, in the paper's own
      terminology.
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The optic nerve may be more sensitive to accumulation of polyprenol and this
      could explain why optic nerve atrophy/hypoplasia are ubiquitous in SRD5A3-CDG
      but not as frequent in other N-linked CDG.
    explanation: >-
      The proposed explanation for the tissue selectivity, offered by the authors as a
      hypothesis ("may be", "could explain") rather than a finding.
    directness: INDIRECT
  - reference: PMID:42472049
    reference_title: "Tissue-specific expression and regulation of congenital disorders of glycosylation genes: A GTEx-based in silico study."
    supports: REFUTE
    evidence_source: COMPUTATIONAL
    snippet: >-
      Tissues frequently affected in the corresponding disorders did not consistently
      display the highest baseline gene expression, underscoring that higher gene
      expression alone is a poor indicator of tissue susceptibility
    explanation: >-
      Cited as REFUTE against the simplest explanation for why this node exists - that
      the cerebellum and optic pathway are hit because they express SRD5A3 most. Across
      twelve CDG genes in healthy tissue that does not hold. It leaves the selectivity
      unexplained rather than explaining it, which is the point: the hypothesis above,
      that the optic nerve is unusually sensitive to accumulated polyprenol, is not
      competing with a well-supported alternative.
  - reference: PMID:42472049
    reference_title: "Tissue-specific expression and regulation of congenital disorders of glycosylation genes: A GTEx-based in silico study."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      identified tissue-specific deviations from balanced biallelic expression for
      several genes, most notably SRD5A3
    explanation: >-
      SRD5A3 is named as one of the genes showing the strongest tissue-specific allelic
      imbalance, which is a candidate mechanism for tissue selectivity that does not
      depend on total expression level. Graded COMPUTATIONAL: this is an in silico
      analysis of healthy donor tissue, not a measurement in patients.
    directness: INDIRECT
- name: Progressive Ocular and Skeletal Degeneration
  biological_scale: TISSUE
  description: >-
    A second arm of the phenotype: features that worsen with time rather than being
    fixed at birth. Adults with SRD5A3-CDG develop cataracts, retinal degeneration and
    kyphosis that were not present in childhood, so the disorder is progressive rather
    than purely static. This is the basis for the recommendation that children diagnosed
    with SRD5A3-CDG be monitored for these complications as they age.

    The retinal component needs a distinction the sources do not always make. It is
    progressive, but it is not late-onset: cohorts ascertained through eye clinics find
    visual loss by age three and a rod-cone dystrophy on imaging in patients whose
    retinas were never examined that way before. What arrives late is the diagnosis, not
    the disease. Cataract and kyphosis, by contrast, are documented as genuinely absent
    in childhood and present in the same patients as adults.
  downstream:
  - target: Cataract
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Lens opacity developing after childhood in reported adult patients.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        the features that may develop over time with this disorder including kyphosis,
        retinitis pigmentosa, and cataracts
      explanation: >-
        Names the three late-onset features this node groups.
  - target: Retinal dystrophy
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Progressive rod-cone degeneration. Reported from the second decade in
      metabolically ascertained patients and from infancy in ophthalmically ascertained
      ones; the edge is kept on this node because the degeneration itself progresses,
      not because it starts late.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        later in life development of retinal abnormalities including retinitis
        pigmentosa as well as glaucoma and cataracts
      explanation: >-
        The source of the late-onset framing, from a cohort followed by metabolic
        medicine.
    - reference: PMID:28253385
      reference_title: Association of Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation With Early-Onset Retinal Dystrophy.
      supports: REFUTE
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Its retinal phenotype is not well described but could be important for disease
        recognition because it appears to be a consistent primary presenting feature.
      explanation: >-
        Cited as REFUTE against the reading that the retinal degeneration is a
        late-appearing complication. In a series ascertained the other way round it was
        the first thing anyone noticed, which is the opposite claim.
  - target: Nyctalopia
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Night blindness, the symptomatic counterpart of the rod dysfunction found on
      electrodiagnostic testing.
    evidence:
    - reference: PMID:28253385
      reference_title: Association of Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation With Early-Onset Retinal Dystrophy.
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        electrodiagnostic testing revealed rod and cone dysfunction in the 5 patients
        tested
      explanation: >-
        The measured rod dysfunction that the reported night blindness corresponds to.
        Quoted rather than the symptom list itself because that sentence is broken up by
        bracketed measurement values.
  - target: Myopia
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Short sight, reported alongside the retinal dystrophy and in some patients marked.
    evidence:
    - reference: PMID:31638560
      reference_title: "Review of SRD5A3 Disease-Causing Sequence Variants and Ocular Findings in Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation, and a Detailed New Case."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        with a myopic refractive error of -3.0 dioptre sphere
      explanation: >-
        A measured refractive error in a patient with the retinal dystrophy.
  - target: Kyphosis
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Progressive spinal deformity in older patients.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        including kyphosis, retinitis pigmentosa, and cataracts
      explanation: >-
        Kyphosis among the features that develop with age.
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All individuals diagnosed with SRD5A3-CDG as children will need close monitoring
      for the development of cataracts, retinitis pigmentosa, and kyphosis as they age.
    explanation: >-
      The clinical consequence of this node being progressive rather than static.
phenotypes:
- name: Intellectual disability
  category: Clinical
  description: >-
    Cognitive impairment, ranging from mild in some compound heterozygous adults to
    severe with absent speech in patients homozygous for null alleles. Present in
    nearly all reported patients.
  phenotype_term:
    preferred_term: Intellectual disability
    term:
      id: HP:0001249
      label: Intellectual disability
  frequency: VERY_FREQUENT
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      nearly every affected individual reported to date has cognitive delays,
      nystagmus/optic atrophy/optic hypoplasia, and ataxia
    explanation: >-
      "Nearly every affected individual" is the basis for the VERY_FREQUENT band.
- name: Nystagmus
  category: Clinical
  description: >-
    Horizontal nystagmus, frequently noticed in infancy and often the finding that
    first brings the patient to ophthalmology.
  phenotype_term:
    preferred_term: Nystagmus
    term:
      id: HP:0000639
      label: Nystagmus
  frequency: VERY_FREQUENT
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      nearly every affected individual reported to date has cognitive delays,
      nystagmus/optic atrophy/optic hypoplasia, and ataxia
    explanation: >-
      Nystagmus among the near-universal features.
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The clinical features were psychomotor retardation, pathological nystagmus,
      slight muscular hypotonia and microcephaly.
    explanation: >-
      Nystagmus in the index patient of an independent report.
- name: Optic atrophy
  category: Clinical
  description: >-
    Degeneration of the optic nerve, reported alongside and distinct from
    developmental hypoplasia. Its progressive character indicates a defect in nerve
    maintenance rather than only in its formation.
  phenotype_term:
    preferred_term: Optic atrophy
    term:
      id: HP:0000648
      label: Optic atrophy
  frequency: VERY_FREQUENT
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      why optic nerve atrophy/hypoplasia are ubiquitous in SRD5A3-CDG
    explanation: >-
      The authors describe the optic nerve findings as ubiquitous in this disorder,
      which is the basis for the band.
- name: Optic nerve hypoplasia
  category: Clinical
  description: >-
    Developmental underdevelopment of the optic nerve, in some patients severe and
    asymmetric.
  phenotype_term:
    preferred_term: Optic nerve hypoplasia
    term:
      id: HP:0000609
      label: Optic nerve hypoplasia
  frequency: FREQUENT
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In addition to optic nerve hypoplasia/atrophy, multiple individuals have been
      reported with colobomas of the optic nerve and iris
    explanation: >-
      Establishes optic nerve hypoplasia as a recurrent finding. The source groups
      hypoplasia with atrophy rather than counting them separately, so the band here is
      set one step below the combined figure.
- name: Ataxia
  category: Clinical
  description: >-
    Cerebellar ataxia with unstable gait; several patients did not walk until age
    four and remain unsteady.
  phenotype_term:
    preferred_term: Ataxia
    term:
      id: HP:0001251
      label: Ataxia
  frequency: VERY_FREQUENT
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      and ataxia (some with documented vermis hypoplasia)
    explanation: >-
      Ataxia listed among the near-universal features in the same sentence as cognitive
      delay and the optic findings.
- name: Cerebellar vermis hypoplasia
  category: Clinical
  description: >-
    Hypoplasia of the inferior cerebellar vermis on MRI, the structural correlate of
    the ataxia in a proportion of patients. Not universal: some patients with ataxia
    have normal imaging.
  phenotype_term:
    preferred_term: Cerebellar vermis hypoplasia
    term:
      id: HP:0001320
      label: Cerebellar vermis hypoplasia
  frequency: FREQUENT
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Head MRI found a hypoplastic inferior cerebellar vermis.
    explanation: >-
      Documented imaging finding in a reported patient. The qualifier "some with
      documented vermis hypoplasia" elsewhere in the same paper is why this is banded
      below the ataxia itself.
  - reference: PMID:41769439
    reference_title: "Neuro-Ophthalmic Presentation of Steroid 5a-Reductase Type 3 Congenital Disorder of Glycosylation: A Case of Monozygotic Twins."
    supports: REFUTE
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      despite prominent neurological symptoms, brain MRIs at 20 months were entirely
      normal, showing no evidence of cerebellar hypoplasia or atrophy typically
      associated with this condition
    explanation: >-
      Two monozygotic twins with the recurrent nonsense allele, marked hypotonia,
      nystagmus and optic atrophy, and no cerebellar abnormality on imaging at 20
      months. Cited as REFUTE against reading this finding as a constant, and because
      it adds a timing dimension the other reports do not: a normal early scan does not
      exclude the diagnosis.
- name: Hypotonia
  category: Clinical
  description: >-
    Generally mild muscular hypotonia, present from infancy.
  phenotype_term:
    preferred_term: Hypotonia
    term:
      id: HP:0001252
      label: Hypotonia
  frequency: FREQUENT
  evidence:
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      psychomotor retardation, pathological nystagmus, slight muscular hypotonia and
      microcephaly
    explanation: >-
      Hypotonia among the presenting features of the index patient.
- name: Iris coloboma
  category: Clinical
  description: >-
    Structural coloboma of the iris, sometimes with optic nerve coloboma in the same
    patient. A variable feature: siblings with the same homozygous null allele differ
    in whether they have it.
  phenotype_term:
    preferred_term: Iris coloboma
    term:
      id: HP:0000612
      label: Iris coloboma
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      multiple individuals have been reported with colobomas of the optic nerve and
      iris
    explanation: >-
      Establishes coloboma as recurrent but, in contrast to the optic nerve findings,
      not described as ubiquitous.
  - reference: PMID:20700148
    reference_title: Next generation sequencing in a family with autosomal recessive Kahrizi syndrome (OMIM 612713) reveals a homozygous frameshift mutation in SRD5A3.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      a novel syndrome consisting of mental retardation, coloboma, cataract and kyphosis
    explanation: >-
      Coloboma is one of the four defining features of Kahrizi syndrome, which this
      paper mapped to a homozygous SRD5A3 frameshift and showed to be allelic with
      SRD5A3-CDG.
- name: Cataract
  category: Clinical
  description: >-
    Lens opacity, characteristically developing after childhood; in one reported
    family cataracts began at age 17.
  phenotype_term:
    preferred_term: Cataract
    term:
      id: HP:0000518
      label: Cataract
    onset:
      onset_category: ADULT
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      all have severe intellectual disability, cataracts beginning at 17 years and
      kyphosis at 8 years old
    explanation: >-
      Documents the age of cataract onset in a reported adult sibship, which is the
      basis for the onset category.
  - reference: PMID:20700148
    reference_title: Next generation sequencing in a family with autosomal recessive Kahrizi syndrome (OMIM 612713) reveals a homozygous frameshift mutation in SRD5A3.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      a novel syndrome consisting of mental retardation, coloboma, cataract and kyphosis
    explanation: >-
      Cataract is one of the four features of the syndrome that was described
      independently as Kahrizi syndrome before the gene was known, and that this paper
      showed to be allelic with SRD5A3-CDG. It is the reason Kahrizi syndrome is carried
      as a synonym of this entry.
- name: Retinal dystrophy
  category: Clinical
  description: >-
    Progressive retinal degeneration. The age at which it is recognised spans the whole
    reported range and appears to depend on how hard it is looked for rather than on when
    it begins: one patient was found to have it in her twenties, whereas a series
    ascertained through eye clinics rather than through metabolic medicine found visual
    loss by age three in every affected individual, with rod and cone dysfunction on
    electrodiagnostic testing. In a third case, imaging showed retinal dystrophy that had
    not been recognised at all until the study that reported it. Onset is therefore
    curated as childhood, not adult.

    On the binding: only one cited source names the specific entity retinitis pigmentosa,
    in one adult patient. The others report "retinal dystrophy" or "early-onset retinal
    dystrophy" without establishing the rod-before-cone sequence that rod-cone dystrophy
    asserts, and the one electrodiagnostic result describes rod and cone dysfunction
    together. So this is bound to the more general HP:0000556 rather than to HP:0000510,
    which would claim more than the sources carry.
  phenotype_term:
    preferred_term: Retinal dystrophy
    term:
      id: HP:0000556
      label: Retinal dystrophy
    onset:
      onset_category: CHILDHOOD
  frequency: FREQUENT
  evidence:
  - reference: PMID:28253385
    reference_title: Association of Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation With Early-Onset Retinal Dystrophy.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Mutations in the SRD5A3 gene may cause early-onset retinal dystrophy, a previously
      underdescribed feature of the SRD5A3-CDG disorder that is progressive and may lead
      to serious visual impairment.
    explanation: >-
      Seven affected individuals from four unrelated families, all homozygous for the
      same allele, ascertained as retinal dystrophy rather than as a glycosylation
      disorder. This is the basis for raising the frequency band and moving onset
      earlier.
  - reference: PMID:28253385
    reference_title: Association of Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation With Early-Onset Retinal Dystrophy.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      electrodiagnostic testing revealed rod and cone dysfunction in the 5 patients
      tested
    explanation: >-
      Functional confirmation that both photoreceptor classes are involved, which is
      what makes the rod-cone dystrophy term the right binding.
  - reference: PMID:31638560
    reference_title: "Review of SRD5A3 Disease-Causing Sequence Variants and Ocular Findings in Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation, and a Detailed New Case."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      showed clear signs of retinal dystrophy not recognized until our investigation
    explanation: >-
      A patient followed for years as congenital nystagmus and optic neuropathy whose
      retinal dystrophy was only found on dedicated imaging. Direct evidence that the
      occasional banding in earlier reports reflects ascertainment.
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      She had cataracts diagnosed in childhood, and in her 20's was found to have
      retinitis pigmentosa.
    explanation: >-
      A documented case with the age at which the retinal degeneration was identified.
      Retained because it is the late end of the ascertainment range described above.
- name: Nyctalopia
  category: Clinical
  description: >-
    Night blindness from childhood, the symptom corresponding to the rod dysfunction
    demonstrated on electroretinography.
  phenotype_term:
    preferred_term: Nyctalopia
    term:
      id: HP:0000662
      label: Nyctalopia
    onset:
      onset_category: CHILDHOOD
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:28253385
    reference_title: Association of Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation With Early-Onset Retinal Dystrophy.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      electrodiagnostic testing revealed rod and cone dysfunction in the 5 patients
      tested
    explanation: >-
      The rod dysfunction that night blindness reflects, measured in the same series
      that reports the symptom. The sentence listing the symptom itself is interrupted
      by bracketed measurement values, so the electrodiagnostic result is quoted
      instead. The band is OCCASIONAL because the symptom is described for the series as
      a whole without a count.
- name: Myopia
  category: Clinical
  description: >-
    Short sight accompanying the retinal dystrophy, in one series marked.
  phenotype_term:
    preferred_term: Myopia
    term:
      id: HP:0000545
      label: Myopia
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:31638560
    reference_title: "Review of SRD5A3 Disease-Causing Sequence Variants and Ocular Findings in Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation, and a Detailed New Case."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      with a myopic refractive error of -3.0 dioptre sphere
    explanation: >-
      A measured refractive error in a patient with SRD5A3-CDG and retinal dystrophy.
- name: Kyphosis
  category: Clinical
  description: >-
    Spinal kyphosis, sometimes with scoliosis, developing over time; marked in adult
    patients.
  phenotype_term:
    preferred_term: Kyphosis
    term:
      id: HP:0002808
      label: Kyphosis
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      She is legally blind and has developed marked kyphosis and scoliosis as an adult.
    explanation: >-
      Documents kyphosis as an acquired adult feature in a reported patient.
  - reference: PMID:20700148
    reference_title: Next generation sequencing in a family with autosomal recessive Kahrizi syndrome (OMIM 612713) reveals a homozygous frameshift mutation in SRD5A3.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      a novel syndrome consisting of mental retardation, coloboma, cataract and kyphosis
    explanation: >-
      Kyphosis is one of the four defining features of the independently described
      Kahrizi syndrome that this paper showed to be caused by SRD5A3, which is why it is
      curated here rather than treated as an incidental finding.
- name: Ichthyosis
  category: Clinical
  description: >-
    Diffuse ichthyosiform skin change, in reported patients initially misdiagnosed as
    psoriasis. Absent in several patients, including some with the same homozygous
    nonsense allele as affected siblings.
  phenotype_term:
    preferred_term: Ichthyosis
    term:
      id: HP:0008064
      label: Ichthyosis
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      She was diagnosed with psoriasis as a child but recent dermatological evaluation
      is more consistent with diffuse ichthyosiform changes and palmoplantar
      keratoderma.
    explanation: >-
      Documents the skin phenotype and the misdiagnosis that commonly precedes it.
- name: Palmoplantar keratoderma
  category: Clinical
  description: >-
    Thickening of the skin of the palms and soles, part of the same dermatological
    phenotype as the ichthyosiform change.
  phenotype_term:
    preferred_term: Palmoplantar keratoderma
    term:
      id: HP:0000982
      label: Palmoplantar keratoderma
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      a more recent dermatological evaluation is more consistent with diffuse
      palmoplantar keratoderma
    explanation: >-
      Documents palmoplantar keratoderma in a reported sibling.
- name: Seizure
  category: Clinical
  description: >-
    Both febrile and non-febrile seizures are reported, in some patients from early
    infancy and requiring medication.
  phenotype_term:
    preferred_term: Seizure
    term:
      id: HP:0001250
      label: Seizure
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      He started non-febrile seizures at 3 months old controlled with medication.
    explanation: >-
      Documents early-onset epilepsy in a reported patient.
  - reference: PMID:41769439
    reference_title: "Neuro-Ophthalmic Presentation of Steroid 5a-Reductase Type 3 Congenital Disorder of Glycosylation: A Case of Monozygotic Twins."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      both twins exhibited generalized tonic-clonic seizures starting in early infancy -
      a feature less commonly reported in SRD5A3-CDG
    explanation: >-
      A second independent report of infantile-onset seizures, and the source that says
      explicitly that seizures are under-reported for this disorder. That is the reason
      the band is left at OCCASIONAL rather than raised: the authors are describing what
      the literature records, not a counted frequency.
- name: Microcephaly
  category: Clinical
  description: >-
    Reduced head circumference, present in some but not all patients.
  phenotype_term:
    preferred_term: Microcephaly
    term:
      id: HP:0000252
      label: Microcephaly
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      slight muscular hypotonia and microcephaly
    explanation: >-
      Microcephaly among the index patient's presenting features.
- name: Reduced protein C activity
  category: Biochemical
  description: >-
    Reduced protein C activity, one of the coagulation abnormalities attributable to
    hypoglycosylation of serum glycoproteins.
  phenotype_term:
    preferred_term: Reduced protein C activity
    term:
      id: HP:0005543
      label: Reduced protein C activity
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Protein C was 50% (normal range 70-180%), protein S was normal at 83%,
      antithrombin III was 53%
    explanation: >-
      A measured protein C activity below the laboratory reference interval quoted
      alongside it in the same sentence, with protein S normal in the same patient.
- name: Reduced antithrombin III activity
  category: Biochemical
  description: >-
    Reduced antithrombin III activity, documented in affected siblings.
  phenotype_term:
    preferred_term: Reduced antithrombin III activity
    term:
      id: HP:0001976
      label: Reduced antithrombin III activity
  frequency: OCCASIONAL
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      protein S was normal at 83%, antithrombin III was 53%
    explanation: >-
      A measured antithrombin III activity below its reference interval, with protein S
      normal in the same patient, so the deficiency is selective rather than a global
      loss of plasma protein.
- name: Abnormal serum transferrin glycoform pattern
  category: Biochemical
  description: >-
    Carbohydrate-deficient transferrin testing shows a type 1 pattern, with raised
    mono-oligo/di-oligo and a-oligo/di-oligo ratios, reflecting whole missing glycans
    rather than truncated ones. This is the screening abnormality that places the
    disorder in CDG type I, but it is not a reliable diagnostic: the hypoglycosylation
    is partial, and one patient with a complete enzyme null still glycosylated about
    70% of their transferrin normally.
  phenotype_term:
    preferred_term: Abnormal isoelectric focusing of serum transferrin
    term:
      id: HP:0003160
      label: Abnormal isoelectric focusing of serum transferrin
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Carbohydrate deficient transferrin testing found: mono-oligo/di-oligo ratio of
      0.44
    explanation: >-
      A measured abnormal transferrin glycoform ratio in a reported patient, with the
      normal cut-off quoted immediately after it in the source.
  - reference: PMID:41732066
    reference_title: Extensive Hypoglycosylation of Serum N-Glycoproteins in SRD5A3 Deficiency.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      the current diagnosis of SRD5A3-CDG by screening assays remains challenging
    explanation: >-
      States that the existing screening assays are not adequate, which is the
      practical consequence of the partial hypoglycosylation.
genetic:
- name: SRD5A3 pathogenic variants
  gene_term:
    preferred_term: SRD5A3
    term:
      id: hgnc:25812
      label: SRD5A3
  association: Causative
  relationship_type: CAUSATIVE
  notes: >-
    All reported alleles act by loss of function. Most are nonsense or frameshift changes in the homozygous state,
    arising in consanguineous families; documented examples include c.57G>A
    (p.Trp19X) and c.603G>A (p.Trp201X). Compound heterozygosity with a missense
    allele is less common and was first described in adults whose cognitive phenotype
    was milder, consistent with residual catalytic activity. Genotype does not
    determine the full phenotype: siblings homozygous for the same nonsense allele
    differ in whether they have structural eye anomalies or skin involvement.
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Sanger sequencing of SRD5A3 was performed and confirmed a novel homozygous
      nonsense mutation
    explanation: >-
      A documented homozygous null allele in affected siblings.
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      These two are of interest because they are among the first individuals reported
      with SRD5A3-CDG due to compound heterozygous mutations with one of the mutations
      being a missense mutation.
    explanation: >-
      Establishes compound heterozygosity with a missense allele as the less common
      genotype, in the two patients with the mildest cognitive phenotype.
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Patient 1 has more severe skin findings and marked short stature while her
      brother has an average stature and more severe structural eye abnormalities with
      retinal and iris colobomas.
    explanation: >-
      Within-family divergence on an identical genotype, which is the evidence that
      genotype does not fix the phenotype here.
  - reference: PMID:35339718
    reference_title: "SRD5A3-CDG: Twins with an intragenic tandem duplication."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Only when applying exome-based copy number analysis, we identified as a second
      compound heterozygous variant a previously not reported tandem duplication of exons
      2-4 in SRD5A3.
    explanation: >-
      A structural allele that standard exome analysis missed, in patients who were
      clinically typical. It is the reason this entry's diagnosis section should not be read
      as saying sequencing alone settles the question.
  - reference: PMID:30019980
    reference_title: Early-onset retinal dystrophy and chronic dermatitis in a girl with an undiagnosed congenital disorder of glycosylation (SRD5A3-CDG).
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Early-onset retinal dystrophy is usually isolated but can also be the presenting
      manifestation of an undiagnosed systemic disease.
    explanation: >-
      Records the presentation route that most often delays diagnosis here: an isolated-looking
      retinal dystrophy in a child who in fact has a multisystem glycosylation disorder.
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      with no functional protein was found in the patient, about 70% of transferrin
    explanation: >-
      Establishes that the patient whose fibroblast lipid measurements underpin the
      mechanistic model carried an allele leaving no functional protein, and that most
      of their transferrin was nonetheless glycosylated normally.
  variants:
  - name: SRD5A3 c.57G>A (p.Trp19X)
    description: >-
      An exon 1 nonsense allele leaving no functional protein. It is the genotype behind
      the entry's central mechanistic observation: fibroblasts from this homozygous patient
      still contained normal dolichol while accumulating polyprenol, which is what forces
      the mechanism onto the polyprenol-to-dolichol ratio rather than onto dolichol
      depletion.

      It is also the most frequently reported allele in this disorder, and the reports
      that carry it are the ones that reframed the eye phenotype: seven individuals from
      four unrelated South Asian families, all homozygous for it, ascertained through eye
      clinics as early-onset retinal dystrophy, and separately a pair of monozygotic
      twins. Whether that reflects a founder haplotype is not established in the cited
      sources — the families are described as unrelated, and no haplotype analysis is
      reported — so it is recorded here as recurrent, not as a founder allele.
    gene:
      preferred_term: SRD5A3
      term:
        id: hgnc:25812
        label: SRD5A3
    type: NONSENSE
    clinical_significance: PATHOGENIC
    functional_effects:
    - function: polyprenol reductase activity
      description: >-
        Abolishes the enzyme. Despite that, dolichol levels are preserved, implying an
        alternative biosynthetic route.
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: IN_VITRO
      snippet: >-
        evaluation of fibroblasts from a child with a homozygous stop codon in exon one at
        p.Trp19X in SRD5A3 found normal levels of dolichol but elevated polyprenol levels
      explanation: >-
        Names the allele and reports the lipid measurements made in cells carrying it.
    - reference: PMID:28253385
      reference_title: Association of Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation With Early-Onset Retinal Dystrophy.
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        identified the same homozygous SRD5A3 c.57G>A, p.(Trp19Ter) variant as the
        underlying cause of early-onset retinal dystrophy in each family
      explanation: >-
        The same allele in four unrelated families, which is the basis for calling it
        recurrent.
    - reference: PMID:41769439
      reference_title: "Neuro-Ophthalmic Presentation of Steroid 5a-Reductase Type 3 Congenital Disorder of Glycosylation: A Case of Monozygotic Twins."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Whole-exome sequencing identified a homozygous nonsense mutation at exon 1
        c.57G>A
      explanation: >-
        An independent report of the same allele, in the twins whose normal early brain
        imaging is cited against the vermis hypoplasia phenotype.
  - name: SRD5A3 c.603G>A (p.Trp201X)
    description: >-
      A nonsense allele found homozygous in two affected siblings. Those siblings differ
      from each other in whether they have structural eye anomalies and in the severity of
      their skin involvement, which is this entry's evidence that genotype does not fix the
      phenotype.
    gene:
      preferred_term: SRD5A3
      term:
        id: hgnc:25812
        label: SRD5A3
    type: NONSENSE
    clinical_significance: PATHOGENIC
    evidence:
    - reference: PMID:27480077
      reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Sanger sequencing of SRD5A3 was performed and confirmed a novel homozygous nonsense
        mutation (c.603G>A, p.Trp201X) in both siblings.
      explanation: >-
        The allele and its homozygous state in both affected siblings.
  - name: SRD5A3 c.460T>C (p.Ser154Pro)
    description: >-
      A missense allele modelled onto the predicted protein structure. It falls in a
      potential active site and is predicted to reduce catalytic efficiency rather than
      abolish it. Missense alleles are the minority in this disorder and are associated
      with the milder end of the cognitive spectrum, which is consistent with residual
      activity.
    gene:
      preferred_term: SRD5A3
      term:
        id: hgnc:25812
        label: SRD5A3
    type: MISSENSE
    clinical_significance: PATHOGENIC
    functional_effects:
    - function: polyprenol reductase activity
      description: >-
        Predicted to reduce catalytic efficiency. This is an in silico prediction from
        structural modelling, not a measured enzyme activity.
    evidence:
    - reference: PMID:33403770
      reference_title: "SRD5A3-CDG: 3D structure modeling, clinical spectrum, and computer-based dysmorphic facial recognition."
      supports: SUPPORT
      evidence_source: COMPUTATIONAL
      snippet: >-
        Herein, we reported a novel SRD5A3 missense pathogenic variant c.460 T > C
        p.(Ser154Pro).
      explanation: >-
        The allele as reported. Its functional consequence is modelled rather than assayed,
        which is why the evidence is graded COMPUTATIONAL.
  - name: SRD5A3 c.203dupC (p.Phe69LeufsX2)
    description: >-
      A frameshift allele found homozygous in a consanguineous family whose condition had
      been described, and mapped, as a separate entity — Kahrizi syndrome, defined by
      intellectual disability with coloboma, cataract and kyphosis. Identifying this
      variant is what showed the two conditions to be the same gene, and is the reason
      Kahrizi syndrome is a synonym of this entry rather than a differential diagnosis.
    gene:
      preferred_term: SRD5A3
      term:
        id: hgnc:25812
        label: SRD5A3
    type: FRAMESHIFT
    clinical_significance: PATHOGENIC
    evidence:
    - reference: PMID:20700148
      reference_title: Next generation sequencing in a family with autosomal recessive Kahrizi syndrome (OMIM 612713) reveals a homozygous frameshift mutation in SRD5A3.
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        identified a homozygous frameshift mutation (c.203dupC; p.Phe69LeufsX2) in the
        gene for steroid 5α-reductase type 3 (SRD5A3) as the disease-causing change in
        this interval
      explanation: >-
        Names the allele and the mapped interval it was found in.
    - reference: PMID:20700148
      reference_title: Next generation sequencing in a family with autosomal recessive Kahrizi syndrome (OMIM 612713) reveals a homozygous frameshift mutation in SRD5A3.
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Our results show that Kahrizi syndrome and this CDG Ix subtype are allelic
        disorders
      explanation: >-
        The conclusion that merges the two named conditions onto one gene, stated by the
        authors who established it.
  - name: SRD5A3 c.436G>A (p.Glu146Lys)
    description: >-
      A missense allele found homozygous in a child followed for years as congenital
      nystagmus with bilateral optic neuropathy, in whom dedicated retinal imaging then
      showed an unrecognised retinal dystrophy. The same patient had optic nerve head
      drusen, which the authors state had not previously been seen in this syndrome; that
      is a single observation and is not curated as a phenotype here.
    gene:
      preferred_term: SRD5A3
      term:
        id: hgnc:25812
        label: SRD5A3
    type: MISSENSE
    clinical_significance: PATHOGENIC
    evidence:
    - reference: PMID:31638560
      reference_title: "Review of SRD5A3 Disease-Causing Sequence Variants and Ocular Findings in Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation, and a Detailed New Case."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        a 12-year-old Czech child harbouring a novel homozygous variant, c.436G>A,
        p.(Glu146Lys) in SRD5A3
      explanation: >-
        The allele and the patient it was found in.
  - name: SRD5A3 c.509A>G (p.Tyr170Cys)
    description: >-
      A missense allele reported in a single patient who had the expected retinal
      dystrophy, ataxia and ichthyosiform skin change, together with a proximal
      limb-girdle pattern of weakness, brisk reflexes and extensor plantar responses. The
      authors propose those last findings as an expansion of the neurological phenotype.
      This entry records the allele but not the proposed new features as phenotypes — see
      the discussion on corticospinal involvement for why.
    gene:
      preferred_term: SRD5A3
      term:
        id: hgnc:25812
        label: SRD5A3
    type: MISSENSE
    clinical_significance: LIKELY_PATHOGENIC
    evidence:
    - reference: PMID:41667393
      reference_title: "Clinical, biochemical and genetic characterization of an Egyptian patient with SRD5A3-congenital glycosylation disorder."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        uncovered a homozygous, likely pathogenic missense variant c.509A > G,
        p.(Tyr170Cys) in SRD5A3 gene
      explanation: >-
        The allele and the significance class the reporting authors assigned, which is
        why this variant is LIKELY_PATHOGENIC rather than PATHOGENIC.
    - reference: PMID:41667393
      reference_title: "Clinical, biochemical and genetic characterization of an Egyptian patient with SRD5A3-congenital glycosylation disorder."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        plasma glycoprotein markers for N- and O-glycosylation showed an aberrant
        glycosylation profile
      explanation: >-
        Biochemical confirmation that the allele produces a glycosylation defect, which
        is what makes it more than a segregating rare variant.
biochemical:
- name: Elevated polyprenol-to-dolichol ratio
  presence: INCREASED
  biomarker_term:
    preferred_term: polyprenol-to-dolichol ratio
  context: >-
    Measured by lipid quantification in cultured patient fibroblasts. NCIT has no term for
    either lipid - searching it returns the reductase protein and the SRD5A3 allele, not
    the analytes - so this biomarker is deliberately left with a free-text preferred_term
    rather than bound to an unrelated code.
  notes: >-
    This is the biochemical signature that defines the disorder and the measurement the
    whole mechanism of this entry rests on. Its diagnostic value is unusual in shape: the
    abnormality is the ratio, not either quantity on its own, because dolichol itself is
    normal in cells with no functional enzyme. A laboratory measuring dolichol alone would
    report a normal result in a patient with a null allele.

    It is also the one readout shown to be pharmacologically movable, which is why it
    appears again as the target of the atorvastatin entry.
  evidence:
  - reference: PMID:22304929
    reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Quantification of dolichol and unreduced polyprenol in the patient's
      fibroblasts demonstrated a high polyprenol/dolichol ratio with normal amounts
      of dolichol
    explanation: >-
      The measurement itself, including the normal dolichol level that makes the ratio
      rather than the level the informative quantity.
  - reference: PMID:20637498
    reference_title: SRD5A3 is required for converting polyprenol to dolichol and is mutated in a congenital glycosylation disorder.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      The presence of residual dolichol in cells depleted for this enzyme suggests the
      existence of an unexpected alternative pathway for dolichol de novo biosynthesis.
    explanation: >-
      Independent confirmation from the gene-discovery paper that dolichol is not depleted,
      which is what a laboratory interpreting a dolichol level alone would misread.
  readouts:
  - target: Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio
    relationship: READOUT_OF
    direction: POSITIVE
    endpoint_context: DIAGNOSTIC
    interpretation: >-
      The ratio is the direct measurement of the node it reports on, in the tissue the
      measurement was made in.
    evidence:
    - reference: PMID:22304929
      reference_title: "Life with too much polyprenol: polyprenol reductase deficiency."
      supports: SUPPORT
      evidence_source: IN_VITRO
      snippet: >-
        demonstrated a high polyprenol/dolichol ratio with normal amounts of dolichol
      explanation: >-
        The raised ratio in patient cells, which is the node's own defining quantity.
  - target: SRD5A3 Polyprenol Reductase Deficiency
    relationship: PHARMACODYNAMIC_MARKER_OF
    direction: POSITIVE
    endpoint_context: PHARMACODYNAMIC
    interpretation: >-
      Correction of the ratio in patient fibroblasts is the readout the one candidate
      therapy was assessed against, so it doubles as a pharmacodynamic marker for any
      future attempt to treat the enzymatic lesion.
    evidence:
    - reference: PMID:41648237
      reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
      supports: SUPPORT
      evidence_source: IN_VITRO
      snippet: >-
        restored polyprenol-to-dolichol ratios in patient fibroblasts
      explanation: >-
        The ratio used as the drug-response endpoint in patient cells.
animal_models:
- name: C. elegans SRD5A3 W19X model
  species: Caenorhabditis elegans
  genotype: srd-5a3 W19X homozygous (patient nonsense allele knock-in)
  publication: PMID:41648237
  description: >-
    The first disease model for SRD5A3-CDG, and the reason a therapeutic candidate exists
    at all. Worms homozygous for the W19X nonsense allele - the same allele carried by the
    patient whose fibroblast lipid measurements underpin this entry's mechanism - reproduce
    developmental delay, neurological dysfunction and mevalonate pathway dysregulation. It
    was built specifically to be screenable, which a mouse model would not be even if one
    existed.

    The choice of allele matters more than "commonly observed in patients" conveys. W19X is
    the most frequently reported allele in this disorder: it is the homozygous genotype of
    all seven patients in the four-family retinal-dystrophy series and of the reported twin
    pair, as well as of the fibroblast donor. So the model is built on the allele that the
    largest share of published patients actually carry, rather than on a convenient null.
  modeled_mechanisms:
  - target: SRD5A3 Polyprenol Reductase Deficiency
    relationship: RECAPITULATES
    fidelity: MODERATE
    description: >-
      Carries a patient allele and reproduces disease-relevant phenotypes at the organismal
      level, including a mevalonate-pathway readout that connects the lesion to the isoprenoid
      supply upstream of polyprenol.
    limitations: >-
      An invertebrate. It cannot model the features that define the human disease clinically
      - the ocular malformations, the cerebellar vermis hypoplasia, the skin phenotype - so
      its value is as a screening system and a mechanism probe, not as a phenocopy.
    readouts:
    - name: Motility in a high-throughput screen
      target: SRD5A3 Polyprenol Reductase Deficiency
      direction: DECREASED
      interpretation: >-
        The screenable phenotype the drug repurposing screen was run against.
      evidence:
      - reference: PMID:41648237
        reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
        supports: SUPPORT
        evidence_source: MODEL_ORGANISM
        snippet: >-
          Using this model, we conducted a high-throughput motility-based drug repurposing
          screen and identified atorvastatin, an FDA-approved HMG-CoA reductase inhibitor, as
          a repurposing candidate.
        explanation: >-
          Names the readout and what it was used for.
    evidence:
    - reference: PMID:41648237
      reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: >-
        we developed the first C. elegans model of SRD5A3-CDG, harboring the homozygous W19X
        nonsense mutation commonly observed in patients. This model recapitulates
        disease-relevant phenotypes, including developmental delays, neurological
        dysfunction, and mevalonate pathway dysregulation.
      explanation: >-
        The model, its genotype, and the phenotypes it reproduces.
diagnosis:
- name: Transferrin screening followed by SRD5A3 sequencing
  diagnosis_term:
    preferred_term: carbohydrate-deficient transferrin measurement
    term:
      id: NCIT:C101016
      label: Carbohydrate-Deficient Transferrin Measurement
  description: >-
    Suspicion is raised by the combination of developmental delay, nystagmus with an
    optic nerve abnormality and ataxia. Carbohydrate-deficient transferrin testing
    showing a type 1 pattern places the disorder among the CDG type I group, and
    sequencing of SRD5A3 confirms it. The screening step is imperfect, because
    hypoglycosylation is partial even in patients with no functional enzyme, so a
    normal or near-normal transferrin profile does not exclude the diagnosis.
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Despite the normal levels of dolichol, there was hypoglycosylation with only 70%
      of transferrin in serum correctly glycosylated.
    explanation: >-
      Quantifies how mild the screening abnormality can be in a patient with a complete
      enzyme null, which is why the screen is not a rule-out.
  - reference: PMID:41732066
    reference_title: Extensive Hypoglycosylation of Serum N-Glycoproteins in SRD5A3 Deficiency.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Some of these alterations could be further pursued to develop glycopeptide-based
      biomarkers as the current diagnosis of SRD5A3-CDG by screening assays remains
      challenging.
    explanation: >-
      States the diagnostic gap and the proposed direction for closing it.
- name: Dysmorphology-assisted recognition
  description: >-
    Recurrent facial features have been catalogued and used to build an automated
    2D facial-image classifier for SRD5A3-CDG. The reported accuracy is from the
    developers' own validation on published and in-house cases; it has not been
    independently replicated, so it is recorded here as a proposed adjunct rather
    than an established diagnostic test.
  evidence:
  - reference: PMID:33403770
    reference_title: "SRD5A3-CDG: 3D structure modeling, clinical spectrum, and computer-based dysmorphic facial recognition."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      Based on facial digital 2D images, we successfully designed and validated a
      SRD5A3-CDG computer based dysmorphic facial analysis, which achieved 92.5%
      accuracy.
    explanation: >-
      The reported performance of the classifier, from the group that built it.
  - reference: PMID:33403770
    reference_title: "SRD5A3-CDG: 3D structure modeling, clinical spectrum, and computer-based dysmorphic facial recognition."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      some recurrent dysmorphic features such as arched eyebrows, wide eyes, shallow
      nasal bridge, short nose, and large mouth
    explanation: >-
      The dysmorphic features the classifier and clinical recognition both rest on.
- name: Retinal electrophysiology and imaging
  diagnosis_term:
    preferred_term: fundus autofluorescence imaging with electroretinography
    term:
      id: NCIT:C162465
      label: Fundus Autofluorescence Imaging
  description: >-
    The route by which this diagnosis is actually reached in a substantial share of
    patients, and the one the transferrin screen above does not cover. Fundus
    autofluorescence and optical coherence tomography detect the retinal dystrophy, and
    electroretinography characterises it as combined rod and cone dysfunction.

    Two things make this more than a supporting investigation. It has found retinal
    dystrophy in a patient followed for years under a different ophthalmic diagnosis, so a
    normal-looking fundus is not a rule-out; and in one series every patient came to
    attention this way rather than through metabolic medicine. NCIT has no
    electroretinography term reachable from Clinical Intervention or Procedure, so the
    binding names the imaging half and the preferred term carries both.
  evidence:
  - reference: PMID:28253385
    reference_title: Association of Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation With Early-Onset Retinal Dystrophy.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Fundus autofluorescence imaging and optical coherence tomographic scans were
      abnormal in all patients, and electrodiagnostic testing revealed rod and cone
      dysfunction in the 5 patients tested.
    explanation: >-
      The yield of this workup in the series where it was applied systematically.
  - reference: PMID:31638560
    reference_title: "Review of SRD5A3 Disease-Causing Sequence Variants and Ocular Findings in Steroid 5α-Reductase Type 3 Congenital Disorder of Glycosylation, and a Detailed New Case."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Examination by spectral domain optical coherence tomography and fundus
      autofluorescence imaging showed clear signs of retinal dystrophy not recognized
      until our investigation
    explanation: >-
      A retinal dystrophy missed for years and found only when this workup was done,
      which is why it is curated as a diagnostic step rather than as characterisation.
- name: Intragenic copy-number analysis when sequencing is uninformative
  diagnosis_term:
    preferred_term: exome-based copy number analysis
  description: >-
    Sequencing does not always find the second allele. In one reported sibship the second
    variant was an intragenic tandem duplication of exons 2 to 4, invisible to the
    standard exome pipeline and found only when copy-number analysis was applied to the
    exome data. The patients were clinically typical, so nothing about the presentation
    signalled that the first result was incomplete.

    This is recorded as a diagnostic step so that a single heterozygous pathogenic variant
    in a clinically compatible patient is not read as excluding the diagnosis. NCIT has no
    term for this analysis under Clinical Intervention or Procedure, so it is left with a
    free-text preferred term.
  evidence:
  - reference: PMID:35339718
    reference_title: "SRD5A3-CDG: Twins with an intragenic tandem duplication."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Only when applying exome-based copy number analysis, we identified as a second
      compound heterozygous variant a previously not reported tandem duplication of exons
      2-4 in SRD5A3.
    explanation: >-
      The allele and the method that found it, in patients whose sequencing had been
      uninformative.
treatments:
- name: Atorvastatin (experimental, not established therapy)
  description: >-
    The only candidate disease-modifying therapy proposed for this disorder, and it is at
    the preclinical stage. An HMG-CoA reductase inhibitor is a counter-intuitive choice
    here - it restricts the isoprenoid supply upstream of polyprenol - but the logic
    follows from this entry's mechanism: if the pathogenic quantity is the
    polyprenol-to-dolichol ratio rather than the dolichol level, then reducing polyprenol
    production should help even though it reduces flux into the pathway overall. In patient
    fibroblasts it did restore the ratio, and in the worm model it rescued the phenotype.
    Recorded here so the reasoning is visible, explicitly NOT as a treatment in clinical
    use: no patient has been treated, and no trial is reported.
  therapeutic_modality: SMALL_MOLECULE
  treatment_term:
    preferred_term: Pharmacotherapy
    term:
      id: NCIT:C15986
      label: Pharmacotherapy
    therapeutic_agent:
    - preferred_term: atorvastatin
      term:
        id: CHEBI:39548
        label: atorvastatin
  target_mechanisms:
  - target: Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio
    description: >-
      The drug acts on the node this entry treats as the operative lesion, and the reported
      readout is that node's own measurement: the polyprenol-to-dolichol ratio in patient
      fibroblasts.
    evidence:
    - reference: PMID:41648237
      reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
      supports: SUPPORT
      evidence_source: IN_VITRO
      snippet: >-
        restored polyprenol-to-dolichol ratios in patient fibroblasts
      explanation: >-
        The correction of the ratio in patient cells, which is what makes the drug act on
        this node. The strongest independent support for the ratio model in this entry,
        because it shows the ratio is pharmacologically movable.
    - reference: PMID:41648237
      reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: >-
        Atorvastatin rescued disease-relevant phenotypes in the worm model
      explanation: >-
        The organismal half of the same result: moving the ratio changes phenotype. Split
        from the fibroblast clause because the two halves of that sentence report
        different kinds of study.
  evidence:
  - reference: PMID:41648237
    reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Atorvastatin rescued disease-relevant phenotypes in the worm model
    explanation: >-
      The whole-organism half of the preclinical result. No human administration is
      reported, which is why this entry names itself experimental.
  - reference: PMID:41648237
    reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      restored polyprenol-to-dolichol ratios in patient fibroblasts
    explanation: >-
      The cell-culture half of the same result. Kept as a separate item because
      evidence_source classifies the study, and one sentence here reports two.
  - reference: PMID:41648237
    reference_title: Repurposing the HMG-CoA Reductase Inhibitor Atorvastatin for SRD5A3-CDG.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Approximately 60 cases have been reported, with treatment limited to symptomatic
      management.
    explanation: >-
      States the current standard of care against which this candidate is proposed:
      symptomatic management only.
- name: Ophthalmological surveillance and management
  description: >-
    No disease-modifying therapy exists. Because cataract, retinitis pigmentosa and
    glaucoma develop over time in patients diagnosed as children, lifelong
    ophthalmological monitoring is recommended, with intervention for the treatable
    components.
  therapeutic_modality: OTHER
  treatment_term:
    preferred_term: Supportive Care
    term:
      id: NCIT:C15747
      label: Supportive Care
  target_phenotypes:
  - preferred_term: Cataract
    term:
      id: HP:0000518
      label: Cataract
  - preferred_term: Retinal dystrophy
    term:
      id: HP:0000556
      label: Retinal dystrophy
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All individuals diagnosed with SRD5A3-CDG as children will need close monitoring
      for the development of cataracts, retinitis pigmentosa, and kyphosis as they age.
    explanation: >-
      The explicit surveillance recommendation this entry implements.
- name: Developmental and neurological supportive care
  description: >-
    Management of the developmental delay, ataxia and epilepsy: anticonvulsants where
    seizures occur, feeding support where oral intake is inadequate, and
    developmental therapies.

    Physical, occupational and speech therapy and orthopaedic management of the kyphosis
    are all reasonable and are named in review prose, but no cited source documents them
    in an SRD5A3-CDG patient, so they are not modelled as separate treatment entries here.
    That is the position for the CDGs generally rather than a gap peculiar to this
    disorder: symptomatic management is what most of them have, and the exceptions are
    the few with a dietary sugar therapy.
  therapeutic_modality: OTHER
  treatment_term:
    preferred_term: Supportive Care
    term:
      id: NCIT:C15747
      label: Supportive Care
  target_phenotypes:
  - preferred_term: Seizure
    term:
      id: HP:0001250
      label: Seizure
  - preferred_term: Ataxia
    term:
      id: HP:0001251
      label: Ataxia
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      He had difficulty taking sufficient calories orally and a G-tube was placed
      providing all nutrition.
    explanation: >-
      Documents the supportive interventions actually used in a reported patient.
  - reference: PMID:39236565
    reference_title: "Treatment of congenital disorders of glycosylation: An overview."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Mostly, we are only able to manage the disease symptoms rather than to address the
      underlying cause.
    explanation: >-
      Places this entry's supportive-only treatment section in its context across the
      CDGs. Graded OTHER because the source is a treatment overview rather than a study
      reporting data.
differential_diagnoses:
- name: DOLK-congenital disorder of glycosylation
  description: >-
    The disorder of the immediately adjacent step, in which dolichol is converted to
    dolichol phosphate. It shares the ichthyosiform skin change, seizures and
    hypotonia but characteristically causes cardiomyopathy and does not produce the
    prominent eye involvement that defines SRD5A3-CDG. The contrast is mechanistically
    informative: it is what makes the ocular phenotype look like a consequence of
    polyprenol accumulation rather than of downstream hypoglycosylation alone.
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      the conversion of dolichol to dolichol phosphate by DOLK leads to cardiomyopathy,
      seizures, hypotonia, and ichthyosiform skin changes, but significant eye
      abnormalities have not been reported
    explanation: >-
      The direct comparison between the two adjacent pathway defects, including the
      absence of eye involvement in the downstream one.
discussions:
- discussion_id: tissue_selectivity_gap
  kind: KNOWLEDGE_GAP
  status: OPEN
  prompt: >-
    Why do the cerebellum and optic pathway bear the brunt of a defect in a
    biosynthetic step that is required by every cell?
  attaches_to:
  - pathophysiology#Cerebellar and Optic Pathway Vulnerability
  - pathophysiology#Systemic Protein Hypoglycosylation
  rationale: >-
    Serum glycoproteomics shows hypoglycosylation is broad, affecting most measured
    glycopeptides, yet the clinical phenotype is dominated by two tissues. One
    hypothesis has been offered - particular sensitivity of the optic nerve to
    accumulated polyprenol, supported by the contrast with DOLK-CDG, which sits one
    step downstream, shares the skin and seizure phenotype and lacks the eye
    involvement. That contrast is suggestive but is a comparison between two small
    case literatures, not an experiment. Nothing in the cited work measures polyprenol
    in neural or retinal tissue, so the edge from hypoglycosylation to tissue
    selectivity is left with unknown intermediates in this pathograph.

    The obvious rival explanation has been checked and does not work. If affected tissues
    were simply the ones that need the most SRD5A3, baseline expression would track the
    phenotype; across twelve CDG genes in healthy human tissue it does not. What that
    survey does find for SRD5A3 is among the strongest tissue-specific departures from
    balanced biallelic expression of any gene it examined, which is a different kind of
    candidate and an untested one. So the gap is not that nobody has looked - it is that
    the cheapest explanation has been ruled out and the remaining ones are unmeasured in
    the tissues that matter.
  evidence:
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The optic nerve may be more sensitive to accumulation of polyprenol and this
      could explain why optic nerve atrophy/hypoplasia are ubiquitous in SRD5A3-CDG but
      not as frequent in other N-linked CDG.
    explanation: >-
      The single proposed explanation for the selectivity, stated as a hypothesis by
      its authors.
  - reference: PMID:41732066
    reference_title: Extensive Hypoglycosylation of Serum N-Glycoproteins in SRD5A3 Deficiency.
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      with 245 of 291 altered glycopeptides decreased in SRD5A3-CDG
    explanation: >-
      Establishes that the biochemical defect is broad, which is what makes the narrow
      clinical selectivity require an explanation.
  - reference: PMID:42472049
    reference_title: "Tissue-specific expression and regulation of congenital disorders of glycosylation genes: A GTEx-based in silico study."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      the molecular basis for their tissue-specific manifestations remains poorly
      understood
    explanation: >-
      Confirms that this gap is recognised across the CDGs rather than being specific to
      how this entry is curated.
- discussion_id: no_mouse_model
  kind: KNOWLEDGE_GAP
  status: OPEN
  prompt: >-
    Why is there still no isolated Srd5a3 mouse model of this disease, and does the one
    published mouse deletion covering the locus tell us anything about it?
  attaches_to:
  - pathophysiology#SRD5A3 Polyprenol Reductase Deficiency
  rationale: >-
    The absence is documentable rather than merely apparent, which is unusual for a
    negative. A spontaneous mouse mutant carrying a megabase-scale 5qC3.3 deletion does
    remove Srd5a3, and homozygotes die at peri-implantation - but that lethality was traced
    to Exoc1, one of the other eight genes in the interval, and confirmed by making an
    Exoc1 knockout separately and by showing that a CRISPR-engineered deletion of the same
    interval lacking the Exoc1 defect does not kill the embryo. So the only mouse in this
    literature that lacks Srd5a3 tells us nothing about Srd5a3, and it would be a mistake to
    read peri-implantation lethality as this gene's phenotype. The gap that leaves is
    consequential: the disorder's defining features are ocular malformation and cerebellar
    hypoplasia, and neither the patient fibroblasts nor the C. elegans model can address
    them. Every tissue-selectivity question in this entry is unanswerable without a
    vertebrate model.
  evidence:
  - reference: PMID:26346620
    reference_title: Peri-implantation lethality in mice carrying megabase-scale deletion on 5qc3.3 is caused by Exoc1 null mutation.
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Genetic investigation revealed deletion of a > 1.2-Mb genomic region containing nine
      genes (Kit, Kdr, Srd5a3, Tmeme165, Clock, Pdcl2, Nmu, Exoc1, and Cep135).
    explanation: >-
      Establishes that the only published mouse lacking Srd5a3 lacks eight other genes too.
  - reference: PMID:26346620
    reference_title: Peri-implantation lethality in mice carrying megabase-scale deletion on 5qc3.3 is caused by Exoc1 null mutation.
    supports: REFUTE
    evidence_source: MODEL_ORGANISM
    snippet: >-
      We concluded that peri-implantation lethality in Kit(WS/WS) was caused by a monogenic
      defect of Exoc1.
    explanation: >-
      Graded REFUTE because it contradicts the reading this mouse would otherwise invite -
      that deleting Srd5a3 is embryonic lethal in mouse. The lethality is Exoc1's.
- discussion_id: alternative_dolichol_pathway
  kind: OPEN_QUESTION
  status: OPEN
  prompt: >-
    What is the alternative route to dolichol that maintains near-normal dolichol
    levels in cells with no functional polyprenol reductase, and does variation in it
    contribute to the phenotypic variability between siblings with identical alleles?
  attaches_to:
  - pathophysiology#Polyprenol Accumulation with Raised Polyprenol-to-Dolichol Ratio
  - mechanistic_hypotheses#polyprenol_competition_model
  rationale: >-
    Two independent groups found residual or normal dolichol in cells lacking the
    enzyme and concluded that another biosynthetic route must exist; neither
    identified it. This is not a loose end but the load-bearing assumption of the
    ratio model, since that model requires dolichol to be present in order for
    polyprenol to compete with it. It also offers a candidate explanation for the
    within-family variability seen on identical homozygous alleles, if tissues differ
    in how far they rely on the alternative route.

    One candidate can now be excluded. Dolichol biosynthesis is no longer thought to be
    the single step SRD5A3 was originally assigned: DHRSX, found in CDG patients after
    SRD5A3 was, defines a three-step detour pathway that has since been shown to be
    conserved in yeast. That makes DHRSX the obvious suspect for the compensating route.
    It is not: in the yeast system, deleting the DHRSX counterpart produces glycosylation
    defects, low dolichol and polyprenol accumulation - the same picture as this disease -
    and expressing DHRSX rescues it while expressing SRD5A3 does not. The two enzymes are
    non-redundant steps in one pathway, so DHRSX cannot substitute for a missing SRD5A3.
    The question stands, with one fewer answer available.
  evidence:
  - reference: PMID:20637498
    reference_title: SRD5A3 is required for converting polyprenol to dolichol and is mutated in a congenital glycosylation disorder.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      The presence of residual dolichol in cells depleted for this enzyme suggests the
      existence of an unexpected alternative pathway for dolichol de novo biosynthesis.
    explanation: >-
      The original observation and its interpretation, from the paper that identified
      the gene.
  - reference: PMID:27480077
    reference_title: "SRD5A3-CDG: Expanding the phenotype of a congenital disorder of glycosylation with emphasis on adult onset features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Different cell lines may also variably rely on the de novo pathway versus
      potential alternate pathways or salvage from turnover.
    explanation: >-
      Raises the tissue-variable reliance that this question asks about.
  - reference: PMID:42201967
    reference_title: The revised three-step detour pathway in dolichol biosynthesis is evolutionarily conserved in budding yeast.
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Deletion of TDA5 caused glycosylation defects, reduced dolichol levels, and
      accumulated polyprenol.
    explanation: >-
      Shows that losing the DHRSX step reproduces this disease's biochemistry in yeast,
      which is what makes it a plausible candidate for the compensating route before the
      rescue experiment rules it out.
  - reference: PMID:42201967
    reference_title: The revised three-step detour pathway in dolichol biosynthesis is evolutionarily conserved in budding yeast.
    supports: REFUTE
    evidence_source: IN_VITRO
    snippet: >-
      All these phenotypes were rescued by expression of DHRSX, but not by DFG10 or
      SRD5A3.
    explanation: >-
      Cited as REFUTE against DHRSX being the alternative route that preserves dolichol
      when SRD5A3 is absent. The two enzymes cannot substitute for each other in either
      direction, which is what excludes the candidate.
- discussion_id: corticospinal_involvement
  kind: KNOWLEDGE_GAP
  status: OPEN
  prompt: >-
    Is there a corticospinal or neuromuscular component to SRD5A3-CDG beyond the
    cerebellar and ocular involvement that defines it?
  attaches_to:
  - pathophysiology#Cerebellar and Optic Pathway Vulnerability
  - genetic#SRD5A3 pathogenic variants
  rationale: >-
    One patient has been reported with a proximal limb-girdle pattern of weakness,
    diffusely brisk reflexes and bilateral extensor plantar responses alongside the usual
    retinal dystrophy, ataxia and ichthyosiform skin change, and the authors propose this
    as an expansion of the neurological phenotype.

    This entry records the allele but does not curate limb-girdle weakness, hyperreflexia
    or extensor plantar responses as phenotypes. That is a deliberate line rather than an
    omission. It is a single patient; upper and lower motor neuron signs in a child with
    a multisystem disorder and marked hypotonia have several possible readings; and the
    features that would distinguish a genuine corticospinal component from an incidental
    one - imaging of the tracts, nerve conduction, muscle histology - are not reported.
    Curating three new phenotypes off one case would put them on the same footing as
    findings established across cohorts.

    Resolving it needs targeted neurological examination in an existing SRD5A3-CDG
    cohort, not another case report.
  evidence:
  - reference: PMID:41667393
    reference_title: "Clinical, biochemical and genetic characterization of an Egyptian patient with SRD5A3-congenital glycosylation disorder."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      neurological assessment revealed proximal limb-girdle pattern of weakness,
      hyperactive reflexes, extensor plantar responses with evidence of cerebellar
      dysfunction
    explanation: >-
      The findings themselves, in the one patient they are reported in.
  - reference: PMID:41667393
    reference_title: "Clinical, biochemical and genetic characterization of an Egyptian patient with SRD5A3-congenital glycosylation disorder."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      We are expanding the neurophenotypic spectrum by reporting proximal limb-girdle
      pattern of weakness combined with diffusely brisk reflexes and bilateral extensor
      plantar responses suggestive of corticospinal or neuromuscular axis involvement
    explanation: >-
      The authors' own framing of the claim as a proposed spectrum expansion, which is
      what this discussion holds open rather than adopting.
- discussion_id: dolichol_supplementation
  kind: OPEN_QUESTION
  status: OPEN
  prompt: >-
    Could supplying dolichol itself bypass the enzymatic block, and is there any
    experimental support for that outside human cells?
  attaches_to:
  - pathophysiology#SRD5A3 Polyprenol Reductase Deficiency
  - treatments#Atorvastatin (experimental, not established therapy)
  rationale: >-
    The one therapeutic candidate this entry carries, atorvastatin, works upstream: it
    was found in a motility screen and acts on the isoprenoid supply, not on the missing
    reduction step. A more direct idea is to replace the product. There is no human
    evidence for it, but there is a plant result that bears on whether the logic holds at
    all: in Arabidopsis, the orthologue of SRD5A3 is essential, its loss impairs protein
    glycosylation, and the resulting dolichol shortage is partially corrected by feeding
    dolichol.

    Two things stop that from being a treatment proposal. The rescue is partial and in a
    plant, whose dolichol requirements and uptake are not those of a human; and this
    entry's own mechanism argues the target may be wrong, since patient cells have normal
    dolichol already and the proposed lesion is the polyprenol-to-dolichol ratio. Adding
    dolichol would lower that ratio, so the idea is not incoherent - but it is untested,
    and the paper proposing it does so as a general suggestion for glycosylation
    disorders rather than for this one.
  evidence:
  - reference: PMID:26628744
    reference_title: POLYPRENOL REDUCTASE2 Deficiency Is Lethal in Arabidopsis Due to Male Sterility.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Shortage of dolichol in PPRD2-deficient cells is partially rescued by PPRD1
      overexpression or by supplementation with dolichol.
    explanation: >-
      The rescue result the idea rests on. Graded OTHER because the system is a plant,
      which is neither a model organism in the vertebrate sense used elsewhere in this
      entry nor an in vitro human system.
  - reference: PMID:26628744
    reference_title: POLYPRENOL REDUCTASE2 Deficiency Is Lethal in Arabidopsis Due to Male Sterility.
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Impaired protein glycosylation seems to be the major factor underlying these
      defects
    explanation: >-
      Establishes that the plant orthologue's loss acts through the same downstream
      consequence as the human disease, which is what makes the rescue relevant at all.
notes: >-
  Naming. Current CDG nomenclature is gene-first, so this entry is SRD5A3-CDG. The
  older designation CDG-Iq (and "CDG1q") is retained as a synonym because much of the
  primary literature uses it.

  GeneReviews. There is no GeneReviews chapter for this disorder, so there is no
  `references:` block and nothing to tag. Searched PubMed for `SRD5A3[TI]
  GeneReviews[TI]` (0 hits) and for `SRD5A3 AND GeneReviews[book]`, which returns one
  record: PMID:20301507, "Congenital Disorders of N-Linked Glycosylation and Multiple
  Pathway Overview", whose own title marks it RETIRED CHAPTER, FOR HISTORICAL REFERENCE
  ONLY. Recorded here so the search is not repeated.

  Frequency bands. The largest phenotype review grades features qualitatively -
  "nearly every affected individual", "ubiquitous", "multiple individuals have been
  reported" - and bands here follow that ordering, with each phenotype's evidence
  explanation saying which wording it rests on. Where a feature is documented only in
  individual case descriptions, OCCASIONAL is used as the conservative default rather
  than a band inferred from case counts in a literature of about 43 patients.

  Counted denominators do exist in one place and are not used as frequency bands. The
  four-family retinal-dystrophy series reports 7 of 7 with abnormal retinal imaging, 5 of
  5 tested with rod and cone dysfunction, and 6 of 7 with learning difficulties. Those are
  real numerators, but every patient in that series is homozygous for the same allele and
  every one was ascertained through an eye clinic, so the proportions describe that
  ascertainment rather than the disease. They are quoted in the relevant evidence
  explanations and were the reason the retinal phenotype's band was raised, but no band in
  this entry is set from them arithmetically.

  Mechanism, and what is not asserted. The pathograph deliberately routes the
  phenotype through the polyprenol-to-dolichol ratio rather than through dolichol
  depletion, because two independent groups measured normal dolichol in enzyme-null
  cells. The competition step itself is inferred from that ratio and is recorded as a
  hypothesis with its authors' hedging preserved. The edge from systemic
  hypoglycosylation to the cerebellar and optic phenotype is marked
  INDIRECT_UNKNOWN_INTERMEDIATES rather than being filled in with a plausible route.

  Models, and a documented absence. There is no vertebrate model of this disease, and that
  is a cited finding rather than an assertion: the one published mouse lacking Srd5a3 lacks
  eight neighbouring genes as well, and its peri-implantation lethality was traced to Exoc1
  rather than to Srd5a3. That is recorded as a knowledge gap with REFUTE-graded evidence
  against the obvious misinterpretation. The models that do exist are a C. elegans W19X
  line, curated under `animal_models:` because a worm is an animal, and patient fibroblasts,
  which appear as the setting of in vitro evidence rather than as a model entry.

  Therapy. Atorvastatin is recorded as an explicitly experimental entry, not as treatment
  in use. No patient has received it for this indication and no trial is reported; what
  exists is a rescue in the worm model and correction of the polyprenol-to-dolichol ratio in
  patient fibroblasts, reported in one sentence that this entry splits into a
  MODEL_ORGANISM item and an IN_VITRO one because it carries both. It is included because
  the reasoning is mechanistically informative -
  an HMG-CoA reductase inhibitor helps only if the pathogenic quantity is the ratio rather
  than the dolichol level, so the result is also evidence for this entry's central model.
  No clinical trials are recorded, because none is reported.
datasets: []

📚

References & Deep Research

Deep Research

1

OpenScientist ▸

SRD5A3-Congenital Disorder of Glycosylation (SRD5A3-CDG): A Comprehensive Disease Characterization Report

openscientist-autonomous 27 citations 2026-09-01T16:13:36.656894

SRD5A3-Congenital Disorder of Glycosylation (SRD5A3-CDG): A Comprehensive Disease Characterization Report

Category: Mendelian (autosomal recessive inborn error of metabolism) Compiled: 2026-09-01 | Evidence base: primary literature (PMIDs cited inline)

Summary

SRD5A3-Congenital Disorder of Glycosylation (SRD5A3-CDG; OMIM #612379; CDG type Iq; Orphanet ORPHA:79320) is a rare autosomal recessive inborn error of metabolism caused by biallelic loss-of-function variants in SRD5A3 (steroid 5α-reductase type 3; chromosome 4q12; HGNC:24420; OMIM 611715). The gene encodes polyprenol reductase, the enzyme that catalyzes the final, committed step of dolichol biosynthesis — reduction of the α-isoprene unit of polyprenol to form dolichol (PMID: 20637498). Because dolichol is the obligate lipid carrier for the dolichol-linked oligosaccharide (LLO) precursor of protein N-glycosylation, and also anchors glycans used in O-/C-mannosylation and glycosylphosphatidylinositol (GPI) anchor biosynthesis, enzyme deficiency produces a CDG type I* biochemical defect (hypoglycosylation of nascent glycoproteins) that branches to affect multiple glycosylation pathways at the endoplasmic reticulum (ER) membrane.

Clinically, SRD5A3-CDG is a congenital multisystem neuro-ophthalmologic-dermatologic disorder. Its consistent core features are intellectual disability/psychomotor delay, muscular hypotonia, cerebellar ataxia with cerebellar hypoplasia/atrophy, and a characteristic ocular spectrum (congenital nystagmus, optic disc pallor/optic atrophy, early-onset retinal dystrophy, ocular coloboma, cataract), frequently accompanied by ichthyosiform skin/chronic dermatitis. The historically separate Kahrizi syndrome (OMIM 612713: intellectual disability, coloboma, cataract, kyphosis) is allelic — indeed the same disorder (PMID: 20700148). Adult-onset/progressive features include kyphosis/scoliosis, retinitis pigmentosa, and cataracts (PMID: 27480077).

The disorder is ultrarare, with roughly 60 cases reported worldwide and approximately 23 distinct pathogenic variants described as of 2022. Diagnosis relies on exome/genome sequencing (including intragenic copy-number analysis) supported by a CDG-I serum transferrin pattern and biochemical demonstration of an elevated polyprenol/dolichol ratio. No disease-modifying therapy exists; management is symptomatic and multidisciplinary. Repurposing of the HMG-CoA reductase inhibitor atorvastatin and dietary dolichol supplementation remain experimental. This report synthesizes eight confirmed findings across 35 reviewed papers to populate the disease knowledge-base template.

Key Findings

Finding 1 — Genetic cause: biallelic loss-of-function SRD5A3 variants encoding polyprenol reductase

SRD5A3 (chromosome 4q12; HGNC:24420; OMIM 611715) encodes polyprenol reductase, which reduces the α-isoprene unit of polyprenol to form dolichol. Dolichol is the obligate lipid carrier for the dolichol-linked oligosaccharide precursor used in protein N-glycosylation, O-/C-mannosylation, and GPI-anchor synthesis. Biallelic pathogenic variants cause SRD5A3-CDG (a CDG type I disorder; OMIM #612379), inherited in an autosomal recessive manner. The seminal functional study established both gene function and disease causation: "We found that SRD5A3 is necessary for the reduction of the alpha-isoprene unit of polyprenols to form dolichols, required for synthesis of dolichol-linked monosaccharides, and the oligosaccharide precursor used for N-glycosylation"* (PMID: 20637498).

Notably, the same study observed residual dolichol in enzyme-depleted cells: "The presence of residual dolichol in cells depleted for this enzyme suggests the existence of an unexpected alternative pathway for dolichol de novo biosynthesis" (PMID: 20637498). This alternative/"detour" pathway — recently shown to be evolutionarily conserved in budding yeast (PMID: 42201967) — explains why patients retain partial glycosylation capacity despite null variants and likely accounts for survival compatible with life in this disorder.

Finding 2 — Core clinical phenotype: neurodevelopmental, ophthalmologic, cerebellar, and cutaneous involvement

Across independent patient cohorts, SRD5A3-CDG presents a recognizable multisystem picture: intellectual disability/psychomotor delay, muscular hypotonia, cerebellar ataxia with cerebellar hypoplasia/atrophy, congenital nystagmus, optic disc pallor/optic atrophy, early-onset retinal dystrophy, ocular coloboma, cataract, and ichthyosiform skin/chronic dermatitis. A concise clinical summary describes "a severe metabolic disease manifesting as muscle hypotonia, developmental delay, cerebellar ataxia and ocular symptoms; typically, nystagmus and optic disc pallor" (PMID: 31638560).

The ocular phenotype is a prominent and often presenting feature, with "early-onset retinal dystrophy as a primary manifestation" (PMID: 28253385). The allelic Kahrizi syndrome highlights the coloboma-cataract-kyphosis axis: "a novel syndrome consisting of mental retardation, coloboma, cataract and kyphosis (Kahrizi syndrome, OMIM 612713)" (PMID: 20700148). Phenotypic diversity is real: monozygotic twins have been reported with early-infancy generalized tonic-clonic seizures (a less common feature) yet entirely normal brain MRI at 20 months, demonstrating that the characteristic cerebellar structural anomalies may be absent early in life (PMID: 41769439).

Suggested HPO terms: HP:0001249 (Intellectual disability), HP:0001252 (Hypotonia), HP:0001251 (Ataxia), HP:0001321 (Cerebellar hypoplasia), HP:0000639 (Nystagmus), HP:0000648 (Optic atrophy), HP:0000556 (Retinal dystrophy), HP:0000589 (Coloboma), HP:0000518 (Cataract), HP:0008064 (Ichthyosis), HP:0002650 (Scoliosis), HP:0002808 (Kyphosis), HP:0001250 (Seizure).

Finding 3 — Biochemical signature: extensive serum hypoglycosylation, elevated polyprenol/dolichol ratio, CDG-I transferrin pattern

Patient fibroblasts exhibit a high polyprenol/dolichol ratio with normal dolichol amounts, the biochemical fingerprint of the enzyme defect: "Quantification of dolichol and unreduced polyprenol in the patient's fibroblasts demonstrated a high polyprenol/dolichol ratio with normal amounts of dolichol" (PMID: 22304929). Serum transferrin isoelectric focusing shows a CDG type I pattern, though a caveat exists — up to ~70% of transferrin may be correctly glycosylated in some patients, so screening can be falsely reassuring, and transferrin protein variants can further confound interpretation (as documented in PMM2-CDG, PMID: 37876147).

Quantitative serum N-glycoproteomics reveals the breadth of the defect: "Extensive hypoglycosylation of serum proteins was observed in patients, with 245 of 291 altered glycopeptides decreased in SRD5A3-CDG" (PMID: 41732066), affecting haptoglobin, plasma serine protease inhibitor, alpha-1-B glycoprotein, alpha-2-macroglobulin, ceruloplasmin, and albumin (including at non-canonical sites). Albumin-derived glycopeptides are emerging as diagnostic biomarkers across CDG subtypes including SRD5A3-CDG (PMID: 41713138). Molecular diagnosis is confirmed by exome/genome sequencing with intragenic copy-number analysis, because structural variants occur: "we identified as a second compound heterozygous variant a previously not reported tandem duplication of exons 2-4 in SRD5A3" (PMID: 35339718).

Finding 4 — No disease-modifying therapy; symptomatic management with atorvastatin repurposing under investigation

With ~60 reported cases, treatment is limited to symptomatic/supportive management: developmental therapies, ophthalmologic and orthopedic care, and seizure control. A recent study developed the first high-throughput disease models and reported repurposing of the HMG-CoA reductase inhibitor atorvastatin as an experimental therapeutic strategy: "Approximately 60 cases have been reported, with treatment limited to symptomatic management" (PMID: 41648237). Dietary dolichol supplementation has been proposed conceptually, supported by plant polyprenol-reductase rescue experiments (Finding 5). A broad 2024 CDG treatment overview reinforces that most CDG remain symptomatically managed: "Mostly, we are only able to manage the disease symptoms rather than to address the underlying cause" (PMID: 39236565), while noting that "Innovative therapies, targeting both the root cause and resulting manifestations, have transitioned from the research stage to practical application" for some subtypes (e.g., dietary sugar therapies in MPI-, PGM1-, and PMM2-CDG).

Finding 5 — Evolutionary conservation: plant/yeast models recapitulate the defect and dolichol rescues it

The polyprenol-to-dolichol reduction step is deeply conserved. Arabidopsis thaliana orthologs PPRD1 and PPRD2 encode polyprenol reductases: "which are orthologous to human SRD5A3 (steroid 5α reductase type 3) and encode polyprenol reductases responsible for conversion of polyprenol to dolichol in Arabidopsis thaliana" (PMID: 26628744). PPRD2 deficiency is lethal (male sterility) and is partially rescued by dolichol: "Shortage of dolichol in PPRD2-deficient cells is partially rescued by PPRD1 overexpression or by supplementation with dolichol" (PMID: 26628744), implicating impaired protein glycosylation as the major underlying factor. The conserved dolichol biosynthesis "detour" pathway in budding yeast (PMID: 42201967) provides an additional tractable model system. These conserved systems provide the mechanistic rationale for dietary dolichol supplementation as a conceptual therapeutic avenue.

Finding 6 — Ultrarare disorder with variable expressivity, adult-onset progression, and founder alleles

Fewer than ~60 cases have been reported worldwide, with ~23 distinct variants known by 2022: "So far, only 23 distinct mutations were described" (PMID: 35339718). Inheritance is autosomal recessive with high representation of consanguineous families and population founder alleles (e.g., p.Gln96delinsX in Baluchi/South Asian families). Expressivity is variable even for recurrent alleles: "Homozygosity for the SRDA3 deletion p.Gln96delinsX is not always associated with ocular coloboma" (PMID: 30019980). Comparison of children and adults with SRD5A3 mutations delineated progressive/adult-onset features: "allowing us to delineate the features that may develop over time with this disorder including kyphosis, retinitis pigmentosa, and cataracts" (PMID: 27480077). In the 280-patient FCDGC natural history cohort, dolichol-metabolism disorders (which include SRD5A3-CDG) comprised ~5% of participants (PMID: 38959600).

Finding 7 — Mechanism: dolichol deficiency impairs multiple ER glycosylation pathways

SRD5A3-derived dolichol is required for synthesis of dolichol-linked monosaccharides and the oligosaccharide precursor (Glc3Man9GlcNAc2-PP-dolichol) assembled in the ER during the dolichol cycle: "required for synthesis of dolichol-linked monosaccharides, and the oligosaccharide precursor used for N-glycosylation" (PMID: 20637498). Dolichyl-phosphate (Dol-P) is a rate-limiting intermediate of N-glycosylation and is recycled to the cytoplasmic ER leaflet after cleavage of dolichyl pyrophosphate: "During protein N-glycosylation, dolichyl pyrophosphate (Dol-P-P) is discharged in the lumenal monolayer of the endoplasmic reticulum (ER)" (PMID: 18077451). Because dolichol also anchors glycans in O-/C-mannosylation and GPI-anchor biosynthesis, the deficiency produces a CDG type I biochemical defect (hypoglycosylation of nascent proteins) that branches to impair those pathways as well.

Finding 8 — CDG therapy is largely symptomatic; no established SRD5A3-specific therapy and no isolated mouse disease model

A 2024 overview (Quelhas & Jaeken) confirms that available CDG treatment options remain limited and mostly symptomatic, though targeted root-cause therapies have recently reached practice for some subtypes (PMID: 39236565). For SRD5A3-CDG specifically, no disease-specific therapy is established; atorvastatin repurposing is experimental. A key model-organism gap exists: no isolated Srd5a3-knockout mouse disease model has been characterized — a mouse ~1.2 Mb 5qC3.3 deletion encompassing Srd5a3 caused peri-implantation lethality attributable to Exoc1, not Srd5a3: "deletion of a > 1.2-Mb genomic region containing nine genes (Kit, Kdr, Srd5a3, Tmeme165, Clock, Pdcl2, Nmu, Exoc1, and Cep135)" (PMID: 26346620).

Detailed Section-by-Section Report

1. Disease Information

Overview. SRD5A3-CDG is an autosomal recessive congenital disorder of glycosylation of the dolichol-metabolism subgroup. Dysfunction of polyprenol reductase blocks the terminal step of dolichol synthesis, reducing availability of the LLO precursor required for protein N-glycosylation and impairing other dolichol-dependent glycan pathways, producing a congenital multisystem disorder dominated by neurodevelopmental, cerebellar, ophthalmologic, and cutaneous features.

Key identifiers: - OMIM (disease): #612379 (Congenital disorder of glycosylation, type Iq) - OMIM (allelic): #612713 (Kahrizi syndrome) - OMIM (gene): 611715 (SRD5A3) - Orphanet: ORPHA:79320 (SRD5A3-CDG) - Gene / HGNC: SRD5A3 / HGNC:24420 - Chromosome: 4q12 - Suggested MONDO: MONDO:0012997 (congenital disorder of glycosylation, type Iq) — verify against current MONDO release - ICD-10: E77.8 (other disorders of glycoprotein metabolism); ICD-11: 5C51.1 (disorders of protein glycosylation) — subtype-level codes are not disease-specific - MeSH:* Congenital Disorders of Glycosylation (D018981) — no SRD5A3-specific descriptor

Synonyms / alternative names: SRD5A3-CDG; CDG type Iq (CDG-Iq); steroid 5α-reductase type 3 deficiency; polyprenol reductase deficiency; Kahrizi syndrome (allelic); congenital disorder of glycosylation with intellectual disability, coloboma, cataract and kyphosis.

Information source type. Disease-level knowledge derives predominantly from aggregated disease-level resources (OMIM, Orphanet) and from published individual/small-cohort case reports and case series plus the multicenter FCDGC natural history study — not from large EHR datasets, consistent with ultrarare-disease evidence.

2. Etiology

Causal factors — genetic. SRD5A3-CDG is a monogenic disorder caused by biallelic (homozygous or compound heterozygous) loss-of-function variants in SRD5A3 (PMID: 20637498). There is no environmental or infectious cause.

Genetic risk factors. The only risk factor is inheritance of two pathogenic SRD5A3 alleles. Consanguinity substantially increases risk (many cases arise in consanguineous families), and founder alleles exist in specific populations (e.g., p.Gln96delinsX in Baluchi/South Asian families; PMID: 30019980).

Environmental risk factors / protective factors / gene-environment interactions. Not applicable — as a Mendelian metabolic disorder, no established environmental risk, protective, or gene-environment interaction factors have been reported. No genetic modifier or protective alleles have been defined, though the conserved dolichol "detour"/alternative biosynthesis pathway (PMID: 20637498; PMID: 42201967) likely modulates residual glycosylation capacity and phenotypic severity.

3. Phenotypes

Phenotype	Type	Onset	Frequency	Suggested HPO
Intellectual disability / psychomotor delay	Cognitive/developmental	Congenital/infantile	Very frequent (near-universal)	HP:0001249
Muscular hypotonia	Clinical sign	Neonatal/infantile	Very frequent	HP:0001252
Cerebellar ataxia	Neurological sign	Infantile/childhood	Frequent	HP:0001251
Cerebellar hypoplasia/atrophy	Imaging/structural	Congenital (may be absent early)	Frequent	HP:0001321 / HP:0001272
Congenital nystagmus	Ophthalmologic sign	Congenital/infantile	Frequent	HP:0000639
Optic disc pallor / optic atrophy	Ophthalmologic sign	Infantile	Frequent	HP:0000648
Early-onset retinal dystrophy	Ophthalmologic	Early childhood	Frequent (can be presenting)	HP:0000556
Ocular coloboma	Physical malformation	Congenital	Variable	HP:0000589
Cataract	Ophthalmologic	Congenital→adult	Variable/progressive	HP:0000518
Ichthyosiform skin / chronic dermatitis	Cutaneous	Infantile	Frequent	HP:0008064
Kyphosis / scoliosis	Skeletal	Adult-onset/progressive	Variable	HP:0002808 / HP:0002650
Retinitis pigmentosa	Ophthalmologic	Adult-onset/progressive	Variable	HP:0000510
Seizures (incl. GTCS)	Neurological	Variable (can be early)	Less common	HP:0001250

Onset, severity, progression. Most features are congenital or infantile-onset. Severity is variable (mild to severe), and expressivity varies even for identical genotypes (PMID: 30019980). The neurodevelopmental deficit is generally stable/non-degenerative in its cognitive component, while several features are progressive (kyphosis, retinitis pigmentosa, cataracts; PMID: 27480077). Structural cerebellar findings may be absent in the first years despite prominent neurological symptoms (PMID: 41769439).

Quality-of-life impact. No disease-specific EQ-5D/SF-36/PROMIS data exist for SRD5A3-CDG. By extrapolation from CDG cohorts, the combination of intellectual disability, ataxia, and visual impairment causes substantial dependency and impaired daily functioning; the FCDGC cohort found 100% of participants had developmental differences and high burdens of neurologic, GI/liver, and musculoskeletal involvement (PMID: 38959600).

4. Genetic / Molecular Information

Causal gene. SRD5A3 (HGNC:24420; OMIM *611715; chromosome 4q12), encoding polyprenol reductase (UniProt Q9H8P0). Loss of function is the disease mechanism (PMID: 20637498).

Pathogenic variants. Approximately 23 distinct variants were described by 2022 (PMID: 35339718). Reported variant classes include: - Frameshift (e.g., the homozygous frameshift in the original Kahrizi-syndrome family; PMID: 20700148) - Nonsense (e.g., c.57G>A, p.Trp19Ter, ACMG-pathogenic; PMID: 41769439) - Missense (e.g., c.509A>G, p.Tyr170Cys, likely pathogenic; PMID: 41667393) - In-frame deletion/indel founder allele (p.Gln96delinsX; PMID: 30019980) - Intragenic structural variants — a tandem duplication of exons 2–4 (PMID: 35339718)

ACMG/AMP classification. Reported variants are predominantly pathogenic or likely pathogenic, consistent with the broader FCDGC cohort in which most CDG variants were classified pathogenic/likely pathogenic (PMID: 38959600). Allele frequencies in gnomAD are very low (consistent with an ultrarare recessive disorder); founder alleles are enriched in specific populations.

Origin and functional consequence. Variants are germline; there is no somatic component. Functional consequence is loss of function (reduced/absent polyprenol reductase activity → impaired dolichol synthesis).

Modifier genes / epigenetics / chromosomal abnormalities. No established modifier genes or epigenetic mechanisms have been reported. The conserved alternative dolichol biosynthesis pathway is a candidate biological modifier of residual glycosylation (PMID: 20637498; PMID: 42201967). No characteristic large-scale chromosomal abnormalities are associated, though intragenic copy-number changes must be sought (PMID: 35339718).

5. Environmental Information

Not applicable. SRD5A3-CDG is a purely genetic Mendelian disorder. No environmental toxins, lifestyle factors, or infectious agents cause or trigger the disease. (Of note, the DPMS/dolichol-phosphate-mannose pathway is a host dependency factor for flaviviruses such as dengue/Zika PMID: 31915280, but this concerns viral biology, not SRD5A3-CDG etiology.)

6. Mechanism / Pathophysiology

Ordered causal chain (initiating lesion → clinical manifestation):

Biallelic loss-of-function variants in SRD5A3 reduce/abolish polyprenol reductase activity. (demonstrated; PMID: 20637498)
Loss of enzyme activity prevents reduction of the α-isoprene unit of polyprenol, so polyprenol accumulates and dolichol formation is impaired → elevated polyprenol/dolichol ratio in cells. (demonstrated; PMID: 22304929)
Reduced dolichol/dolichyl-phosphate limits synthesis of dolichol-linked monosaccharides and the LLO precursor Glc3Man9GlcNAc2-PP-dolichol in the ER (Dol-P is rate-limiting). (demonstrated; PMID: 20637498; PMID: 18077451)
Deficient LLO precursor leads to protein N-hypoglycosylation (a CDG type I defect) — extensive across serum glycoproteins. (demonstrated; PMID: 41732066)
Branch 4a: Reduced dolichol-anchored glycan donors also impair O-/C-mannosylation and GPI-anchor biosynthesis. (inferred from shared dolichol dependency; PMID: 20637498; PMID: 40902550)
Widespread hypoglycosylation disrupts glycoprotein folding, stability, trafficking, and function across many tissues. (inferred; general glycoprotein quality-control biology, PMID: 10794707)
Tissue-level dysfunction — most sensitively in developing brain/cerebellum, retina/eye, and skin — results in the clinical phenotype: intellectual disability, hypotonia, cerebellar ataxia/hypoplasia, the ocular spectrum, and ichthyosis. (clinical correlation; PMID: 31638560)
Branch 6a (residual pathway): A conserved alternative dolichol biosynthesis pathway supplies residual dolichol, mitigating severity and explaining survival and partial glycosylation. (inferred/demonstrated in models; PMID: 20637498; PMID: 42201967)

ASCII pathway diagram:

 SRD5A3 biallelic LoF
│ (loss of polyprenol reductase activity)
▼
 Polyprenol NOT reduced ──► ↑ polyprenol/dolichol ratio
│ (dolichol/Dol-P deficient)
▼
 ↓ Dolichol-linked monosaccharides + LLO (Glc3Man9GlcNAc2-PP-Dol)  [ER membrane]
│
├──► ↓ N-glycosylation (CDG type I)  ──► serum protein hypoglycosylation
├──► ↓ O-/C-mannosylation (branch, inferred)
└──► ↓ GPI-anchor synthesis (branch, inferred)
    │
    ▼
Glycoprotein misfolding / dysfunction (multi-tissue)
    │
┌───────────┼───────────────┬─────────────┐
▼           ▼               ▼             ▼
   Brain/cerebellum  Eye/retina    Skin        Skeleton (progressive)
   ID, hypotonia,    nystagmus,    ichthyosis  kyphosis/scoliosis
   ataxia, hypoplasia optic atrophy,
     retinal dystrophy,
     coloboma, cataract

Molecular pathway / cellular process / compartment. The defect is in dolichol biosynthesis feeding the protein N-glycosylation (dolichol) cycle at the ER membrane. Key GO terms: GO:0019408 (dolichol biosynthetic process), GO:0006486 (protein glycosylation), GO:0006487 (protein N-linked glycosylation), GO:0006506 (GPI anchor biosynthetic process), GO:0035269 (protein O-linked mannosylation). Cellular compartment: GO:0005789 (endoplasmic reticulum membrane), GO:0005783 (endoplasmic reticulum). Enzyme activity: polyprenol reductase (EC 1.3.1.94).

Metabolic changes. Isoprenoid/dolichol lipid metabolism is directly disrupted (elevated polyprenol, relatively preserved absolute dolichol via the detour pathway). Immune involvement is not a primary feature of SRD5A3-CDG (unlike immune-relevant CDG such as MOGS-, PGM3-, VPS13B-CDG). Molecular profiling to date is dominated by serum N-glycoproteomics showing broad hypoglycosylation (PMID: 41732066); a GTEx in-silico study noted that tissue vulnerability in CDG does not simply track baseline gene expression (PMID: 42472049).

Cell types (suggested CL terms): cerebellar Purkinje cell (CL:0000121), photoreceptor cell (CL:0000210), keratinocyte (CL:0000312). Suggested CHEBI: polyprenol (CHEBI:26250), dolichol (CHEBI:23514), dolichyl phosphate (Dol-P), dolichyl diphosphate.

7. Anatomical Structures Affected

Primary organs/systems: central nervous system, especially cerebellum (UBERON:0002037) and cerebrum; eye/retina (UBERON:0000970 / retina UBERON:0000966, optic nerve UBERON:0000941); skin (UBERON:0002097).
Secondary/progressive involvement: skeleton/spine (kyphosis, scoliosis; vertebral column UBERON:0002415).
Body systems: nervous, ophthalmic/sensory, integumentary, musculoskeletal.
Tissue/cell level: neural tissue (cerebellar neurons incl. Purkinje cells), retinal photoreceptors, epidermal keratinocytes.
Subcellular level: endoplasmic reticulum membrane (site of the dolichol cycle; GO:0005789) is the primary affected compartment.
Lateralization: manifestations are typically bilateral/generalized (e.g., bilateral nystagmus, bilateral optic atrophy, symmetric cerebellar involvement).

8. Temporal Development

Onset: Congenital / neonatal-infantile, with hypotonia, nystagmus, and developmental delay evident early; onset pattern is chronic/insidious.
Progression: The cognitive deficit is broadly static, while specific features are progressive — kyphosis, retinitis pigmentosa, and cataracts develop or worsen over time (PMID: 27480077). Cerebellar structural changes may appear or become detectable later, being absent on early MRI in some infants (PMID: 41769439).
Course/duration: Chronic, lifelong. No spontaneous remission. Critical period: early childhood neurodevelopment is the key window for supportive/early-intervention therapies; any future disease-modifying therapy would presumably need early initiation.

9. Inheritance and Population

Epidemiology: Ultrarare; ~60 cases reported worldwide; precise prevalence/incidence are unavailable (well under Orphanet's rare-disease threshold). Dolichol-metabolism disorders were ~5% of the 280-patient FCDGC natural history cohort (PMID: 38959600).
Inheritance: Autosomal recessive (PMID: 20637498).
Penetrance: Complete for biallelic LoF (all reported biallelic carriers are affected), with variable expressivity (PMID: 30019980).
Anticipation / germline mosaicism: Not applicable/not reported (non-repeat-expansion disorder).
Founder effects / consanguinity: Documented — p.Gln96delinsX founder allele in Baluchi/South Asian families; frequent consanguinity (PMID: 30019980).
Carrier frequency: Not precisely established; expected very low given rarity.
Sex ratio: ~1:1 (autosomal); no sex predilection. In the FCDGC cohort overall, sexes were roughly balanced (PMID: 38959600).
Geographic distribution: Cases reported across diverse populations (European, South Asian, Middle Eastern, Egyptian, etc.); founder alleles create regional clustering.

10. Diagnostics

Recommended approach. Diagnosis is genetic-first in the modern setting: exome or genome sequencing including intragenic copy-number/structural-variant analysis (PMID: 35339718), supported by biochemical screening.

Laboratory / biochemical tests: - Serum transferrin isoelectric focusing / CDT analysis → CDG type I pattern. Caveat: can be falsely reassuring (substantial correctly glycosylated transferrin in some patients) and confounded by transferrin protein variants (PMID: 37876147). - Polyprenol/dolichol ratio in fibroblasts (elevated) — biochemical confirmation of the enzyme defect (PMID: 22304929). - Serum N-glycoproteomics demonstrating extensive hypoglycosylation (PMID: 41732066); albumin glycopeptides as emerging biomarkers (PMID: 41713138).

Imaging / functional / electrophysiology: Brain MRI (cerebellar hypoplasia/atrophy — may be normal early); ophthalmologic evaluation with fundus photography, autofluorescence, and electroretinogram (ERG) for retinal dystrophy (PMID: 41667393).

Genetic testing modalities: WES and WGS are the highest-yield tools; targeted CDG gene panels including SRD5A3; single-gene SRD5A3 testing (incl. deletion/duplication analysis). Karyotyping, FISH, mitochondrial DNA testing, and repeat-expansion testing are not applicable.

Clinical criteria / differential diagnosis. No formal consensus criteria exist; diagnosis rests on the phenotype + CDG-I biochemistry + biallelic SRD5A3 variants. Differential diagnosis includes other CDG type I subtypes (notably PMM2-CDG, the most common), other dolichol-pathway CDG (DHDDS-, DPM1/3-, SRD5A3-), Leber congenital amaurosis/early-onset retinal dystrophies, and cerebellar hypoplasia syndromes; distinguishing features are the specific combination of retinal dystrophy, coloboma/cataract, ichthyosis, and the elevated polyprenol/dolichol ratio.

Screening. No newborn screening exists for SRD5A3-CDG. Carrier and cascade testing are appropriate in known families.

11. Outcome / Prognosis

Survival/mortality: No formal survival statistics; the disorder is chronic and compatible with survival into adulthood (adult patients are described; PMID: 27480077), consistent with residual glycosylation via the alternative dolichol pathway. Severe early presentations (e.g., seizures) can occur (PMID: 41769439).
Morbidity/disability: Substantial and lifelong — intellectual disability, ataxia, and progressive visual loss drive major functional impairment and dependency.
Complications/progression: Progressive kyphosis/scoliosis, retinitis pigmentosa, and cataracts (PMID: 27480077); seizures in a minority.
Prognostic factors: Genotype (null vs hypomorphic), residual enzyme/dolichol pathway activity, and severity of neurodevelopmental involvement are the main determinants; validated prognostic biomarkers are not established. The Nijmegen Progression CDG Rating Scale (NPCRS) is used across CDG to grade severity (PMID: 38959600).

12. Treatment

No disease-modifying therapy is established. Management is symptomatic and multidisciplinary (PMID: 41648237; PMID: 39236565): - Neurodevelopmental: early intervention, physical/occupational/speech therapy (NCIT: Rehabilitation Therapy). - Ophthalmologic: low-vision support, cataract surgery as indicated (NCIT: Cataract Surgery). - Orthopedic: management of kyphosis/scoliosis. - Seizure control: standard antiseizure medications (e.g., levetiracetam is effective in related CDG; PMID: 37955240) (NCIT: Levetiracetam).

Experimental / investigational: - Atorvastatin repurposing (HMG-CoA reductase inhibitor) — reported as an experimental strategy in newly developed high-throughput SRD5A3-CDG models (PMID: 41648237). NCIT: Atorvastatin Calcium (C29014). - Dietary dolichol supplementation — conceptual, supported by dolichol rescue in plant PPRD2-deficient models (PMID: 26628744). - Cross-CDG precedents for root-cause "sugar" therapies (not yet demonstrated for SRD5A3-CDG): D-galactose in PGM1-CDG (PMID: 41182978), oral mannose in MPI-CDG, and epalrestat (aldose reductase inhibitor) in PMM2-CDG (PMID: 34652821) illustrate the emerging targeted-therapy landscape (PMID: 39236565).

Pharmacogenomics, gene/cell/RNA therapy, immunotherapy, surgery-as-cure: No SRD5A3-specific advanced therapeutics exist; these remain future directions.

13. Prevention

Primary prevention: Not applicable at the population level (genetic disease). Genetic counseling for at-risk families, carrier testing, and reproductive options (prenatal diagnosis, preimplantation genetic testing) are the principal preventive tools, especially in consanguineous families and founder-allele populations.
Secondary prevention: No population newborn screen; cascade testing in known families enables early diagnosis and supportive intervention.
Tertiary prevention: Proactive surveillance and management of progressive complications (vision, spine, seizures) to reduce morbidity.
Immunization / public-health / environmental interventions: Not applicable.

14. Other Species / Natural Disease

Taxonomy / orthologs: The polyprenol→dolichol reduction step is deeply conserved. Human SRD5A3 has orthologs in Arabidopsis thaliana (PPRD1, PPRD2; PMID: 26628744) and in budding yeast (Saccharomyces cerevisiae), where the three-step dolichol "detour" pathway is conserved (PMID: 42201967). Mouse ortholog: Srd5a3 (PMID: 26346620).
Natural/veterinary disease: No naturally occurring SRD5A3-CDG has been reported in companion animals or wildlife (no OMIA entry identified in this investigation).
Comparative biology: The conservation of the enzyme and the dolichol-rescue phenotype (PPRD2-deficient plants rescued by dolichol) indicates a conserved mechanism linking dolichol supply to protein glycosylation across kingdoms (PMID: 26628744).
Transmission / zoonosis: Not applicable.

15. Model Organisms

Model	Type	Utility / recapitulation	Reference
Arabidopsis thaliana pprd2	Plant genetic	Lethal (male sterility); dolichol shortage; rescued by dolichol supplementation — establishes causal role of dolichol/glycosylation	PMID: 26628744
S. cerevisiae (dolichol detour pathway)	Yeast	Confirms conserved alternative dolichol biosynthesis; tractable for pathway dissection	PMID: 42201967
Patient fibroblasts	In vitro (human)	Elevated polyprenol/dolichol ratio; hypoglycosylation — core biochemical model	PMID: 22304929
High-throughput SRD5A3-CDG disease models	In vitro (recent)	Enabled atorvastatin repurposing screen	PMID: 41648237
Mouse Srd5a3	Mammalian	Gap: no isolated Srd5a3-null disease model characterized; a ~1.2 Mb 5qC3.3 deletion including Srd5a3 caused peri-implantation lethality attributable to Exoc1, not Srd5a3	PMID: 26346620

Model limitations. Plant/yeast models capture the enzymatic and glycosylation defect but not the mammalian neuro-ophthalmologic phenotype. The absence of a validated isolated Srd5a3-knockout mouse model is a major gap for preclinical therapeutic testing.

Mechanistic Model / Interpretation

SRD5A3-CDG is best understood as a substrate-supply failure in the dolichol cycle. The single enzymatic lesion (polyprenol reductase deficiency) sits upstream of the entire dolichol-dependent glycosylation machinery. Its immediate, demonstrated consequence is an elevated polyprenol/dolichol ratio — polyprenol accumulates because it cannot be reduced, while absolute dolichol is partly preserved by a conserved alternative/detour biosynthesis pathway. This residual dolichol is mechanistically important: it explains both the survival of patients (versus embryonic lethality expected from total glycosylation failure) and the observation of substantial correctly glycosylated transferrin in some patients.

Downstream, dolichyl-phosphate limitation throttles assembly of the LLO precursor, producing broad protein N-hypoglycosylation (demonstrated by serum glycoproteomics: 245/291 altered glycopeptides decreased). Because dolichol is a shared currency, the defect branches to O-/C-mannosylation and GPI-anchor synthesis (inferred). The tissue selectivity of the clinical phenotype — brain/cerebellum, eye/retina, skin — reflects the particular sensitivity of these developing tissues to glycoprotein dysfunction rather than tissue-specific enzyme expression (GTEx analysis found baseline expression does not predict CDG tissue vulnerability). The temporal profile (congenital onset with later-emerging kyphosis, retinitis pigmentosa, cataracts) suggests both a developmental component (cerebellar hypoplasia, coloboma) and a slowly progressive/degenerative component (retinal, lens, skeletal).

Evidence Base

PMID	Title (abbrev.)	Supports
20637498	SRD5A3 required for polyprenol→dolichol; mutated in CDG	F001, F007 — gene function, disease causation, residual pathway
22304929	Life with too much polyprenol	F003 — elevated polyprenol/dolichol ratio biomarker
41732066	Extensive hypoglycosylation of serum N-glycoproteins	F003 — 245/291 glycopeptides decreased
35339718	SRD5A3-CDG: intragenic tandem duplication	F003, F006 — CNV diagnosis; ~23 variants
20700148	Kahrizi syndrome frameshift in SRD5A3	F002 — allelic Kahrizi phenotype
31638560	Review of SRD5A3 variants and ocular findings	F002 — core clinical features
28253385	SRD5A3-CDG with early-onset retinal dystrophy	F002 — retinal dystrophy as presenting feature
27480077	SRD5A3-CDG adult-onset features	F006 — progressive kyphosis, RP, cataracts
30019980	Undiagnosed SRD5A3-CDG girl	F006 — founder allele, variable expressivity
26628744	Arabidopsis PPRD2 deficiency	F005 — orthology, dolichol rescue
42201967	Dolichol detour pathway conserved in yeast	F001, F005 — conserved alternative pathway
18077451	Recycling of dolichyl monophosphate	F007 — ER dolichol cycle, Dol-P recycling
41648237	Repurposing atorvastatin for SRD5A3-CDG	F004 — ~60 cases; experimental therapy
39236565	Treatment of CDG: an overview	F004, F008 — symptomatic care; emerging targeted therapies
26346620	Peri-implantation lethality on 5qC3.3	F008 — mouse model gap (Exoc1, not Srd5a3)
38959600	FCDGC natural history cohort (n=280)	Epidemiology — dolichol disorders ~5%; NPCRS
41769439	Monozygotic twins, SRD5A3-CDG	Phenotype diversity; normal early MRI; nonsense variant
41667393	Egyptian SRD5A3-CDG patient	Missense p.Tyr170Cys; ERG diagnostics
41713138	Albumin as a glycoprotein biomarker in CDG	Emerging albumin glycopeptide biomarker
37876147	Misleading transferrin variants in CDG	Diagnostic caveat for transferrin screening
34652821	Epalrestat/sorbitol in PMM2-CDG	Cross-CDG targeted-therapy precedent
41182978	D-galactose in PGM1-CDG	Cross-CDG dietary-sugar precedent
40902550	Genetic disorders of dolichol synthesis and utilization	Review — dolichol pathway/CDG classification

Evidence source types: The evidence base is predominantly human clinical (case reports, case series, natural history cohort) and in vitro (patient fibroblasts, glycoproteomics), complemented by model-organism work in plants and yeast. There are no large randomized trials, no validated mammalian in-vivo disease model, and no computational/GWAS studies relevant to this Mendelian disorder.

Limitations and Knowledge Gaps

Small evidence base. With ~60 cases, all clinical claims derive from case reports/series and one natural history cohort; frequencies are qualitative, and formal prevalence/incidence, survival, and QoL metrics are unavailable.
No validated mammalian disease model. The absence of a characterized isolated Srd5a3-knockout mouse (PMID: 26346620) hampers mechanistic and preclinical therapeutic studies.
Branch mechanisms inferred. Impairment of O-/C-mannosylation and GPI-anchor synthesis is inferred from shared dolichol dependency rather than directly demonstrated in SRD5A3-CDG patients.
Diagnostic pitfalls. Transferrin-based screening can be falsely reassuring; intragenic CNVs require dedicated analysis — both risk under-/mis-diagnosis.
No proven therapy. Atorvastatin repurposing and dolichol supplementation are experimental; no clinical efficacy data exist for SRD5A3-CDG.
Genotype–phenotype correlation is weak. Variable expressivity (even for identical alleles) is unexplained; modifier genes and the quantitative contribution of the alternative dolichol pathway are undefined.

Proposed Follow-up Experiments / Actions

Generate a conditional/tissue-specific Srd5a3 mouse model (e.g., neural- and eye-restricted knockouts) to overcome the peri-implantation lethality confounded by Exoc1, enabling phenotype recapitulation and therapeutic testing.
Validate atorvastatin efficacy in patient-derived models (iPSC-derived neurons/organoids, cerebellar and retinal organoids) and, if positive, design an n-of-few clinical trial with glycosylation biomarkers as endpoints (PMID: 41648237).
Test dietary dolichol supplementation rescue in mammalian patient-derived systems, building on plant/yeast rescue data (PMID: 26628744).
Directly quantify O-/C-mannosylation and GPI-anchor defects in patient cells to confirm the inferred mechanistic branches.
Standardize diagnostics: adopt reflex confirmatory testing (polyprenol/dolichol ratio, glycoproteomics, CNV analysis) whenever transferrin screening is normal but clinical suspicion is high; evaluate albumin glycopeptides as a robust biomarker (PMID: 41713138).
Enroll SRD5A3-CDG patients in the FCDGC natural history study to build prospective longitudinal outcome, NPCRS severity, and QoL data (PMID: 38959600).
Map founder alleles and carrier frequencies in high-consanguinity populations to guide carrier screening and reproductive counseling.

Report compiled from 8 confirmed findings across 35 reviewed publications. Evidence is predominantly human-clinical and in-vitro, supplemented by conserved plant/yeast models. All mechanistic and clinical claims are cited to primary literature by PMID; inferred steps are explicitly labeled.

Artifacts

Reference Validation

Checked with linkml-reference-validator 0.2.1.

Outcome	Count
References checked	27
Resolved	27
Unresolved (possible confabulation)	0
Unverifiable	0
References weighed for topical relevance	27
On topic	17
Off topic	0

All extracted references resolved successfully.

Term Validation

Checked with linkml-term-validator 0.4.5, through the ols: adapter.

Outcome	Count
Terms checked	36
Resolved	32
Unresolved (possible confabulation)	0
Obsolete	2
Unverifiable	2
Terms whose name was checked	24
Terms named correctly	7
Terms named as a different term	13
Terms whose name is worth a second look	4

Terms the report names something else

These identifiers resolve, so nothing about them looks wrong, and the ontology calls them something unrelated to what the report calls them. That usually means the identifier is not the one the sentence needs:

HP:0001249 (2 mentions) - the report calls it "Intellectual disability", "Very frequent (near-universal)"; HP calls it Intellectual disability
HP:0001252 (2 mentions) - the report calls it "Hypotonia", "Very frequent"; HP calls it Hypotonia
HP:0001251 (2 mentions) - the report calls it "Ataxia", "Frequent"; HP calls it Ataxia
HP:0000639 (2 mentions) - the report calls it "Nystagmus", "Frequent"; HP calls it Nystagmus
HP:0000648 (2 mentions) - the report calls it "Optic atrophy", "Frequent"; HP calls it Optic atrophy
HP:0000556 (2 mentions) - the report calls it "Retinal dystrophy", "Frequent (can be presenting)"; HP calls it Retinal dystrophy
HP:0000589 (2 mentions) - the report calls it "Coloboma", "Variable"; HP calls it Coloboma
HP:0000518 (2 mentions) - the report calls it "Cataract", "Variable/progressive"; HP calls it Cataract
HP:0008064 (2 mentions) - the report calls it "Ichthyosis", "Frequent"; HP calls it Ichthyosis
HP:0001250 (2 mentions) - the report calls it "Seizure", "Less common"; HP calls it Seizure
MONDO:0012997 (1 mention) - the report calls it "congenital disorder of glycosylation, type Iq"; MONDO calls it cholestasis-pigmentary retinopathy-cleft palate syndrome
HP:0000510 (1 mention) - the report calls it "Variable"; HP calls it Rod-cone dystrophy
UBERON:0002037 (1 mention) - the report calls it "cerebellum", "Primary organs/systems: central nervous system, especially cerebellum"; UBERON calls it cerebellum

Obsolete terms

These terms are real but deprecated. Citing one is not a fabrication; it does mean the report is naming something the ontology has retired:

GO:0019408 (obsolete dolichol biosynthetic process) (1 mention) - replaced by GO:0043048
GO:0006486 (obsolete protein glycosylation) (1 mention) - replaced by GO:0009101

Terms whose name is worth a second look

The report's name for these is recognisably related to the term's own name without being one of them. A loose paraphrase reads the same way as a citation of the wrong sibling term - and so does a related synonym, which the ontology records precisely because it names something adjacent rather than the same thing - so these are listed rather than judged:

GO:0019408 (1 mention) - the report calls it "dolichol biosynthetic process"; GO calls it obsolete dolichol biosynthetic process
GO:0006486 (1 mention) - the report calls it "protein glycosylation"; GO calls it obsolete protein glycosylation
GO:0035269 (1 mention) - the report calls it "protein O-linked mannosylation"; GO calls it protein O-linked glycosylation via mannose, and lists "protein O-linked mannosylation" among its other names
UBERON:0002097 (1 mention) - the report calls it "skin"; UBERON calls it skin of body, and lists "skin" among its other names

Terms named inconsistently

The report gives these identifiers more than one name of its own:

HP:0001249 - called "Intellectual disability", "Very frequent (near-universal)"
HP:0001252 - called "Hypotonia", "Very frequent"
HP:0001251 - called "Ataxia", "Frequent"
HP:0000639 - called "Nystagmus", "Frequent"
HP:0000648 - called "Optic atrophy", "Frequent"
HP:0000556 - called "Retinal dystrophy", "Frequent (can be presenting)"
HP:0000589 - called "Coloboma", "Variable"
HP:0000518 - called "Cataract", "Variable/progressive"
HP:0008064 - called "Ichthyosis", "Frequent"
HP:0001250 - called "Seizure", "Less common"
UBERON:0002037 - called "cerebellum", "Primary organs/systems: central nervous system, especially cerebellum"

Prefixes with no resolver

Terms carrying these prefixes were not checked either way, because no configured ontology covers them. An unrecognised prefix may name an ontology this run could not reach as easily as one that does not exist, so nothing here is evidence of fabrication: ORPHA.

Ask OpenScientist

OpenScientist Report

Mappings

Inheritance

Mechanistic Hypotheses

Discussions and Knowledge Gaps

Pathophysiology

Pathograph

Phenotypes

Genetic Associations

Medical Actions

Biochemical Markers

Diagnosis

Prevalence

Differential Diagnoses

Animal Models

Source YAML

References & Deep Research

Deep Research

SRD5A3-Congenital Disorder of Glycosylation (SRD5A3-CDG): A Comprehensive Disease Characterization Report

Summary

Key Findings

Finding 1 — Genetic cause: biallelic loss-of-function SRD5A3 variants encoding polyprenol reductase

Finding 2 — Core clinical phenotype: neurodevelopmental, ophthalmologic, cerebellar, and cutaneous involvement

Finding 3 — Biochemical signature: extensive serum hypoglycosylation, elevated polyprenol/dolichol ratio, CDG-I transferrin pattern

Finding 4 — No disease-modifying therapy; symptomatic management with atorvastatin repurposing under investigation

Finding 5 — Evolutionary conservation: plant/yeast models recapitulate the defect and dolichol rescues it

Finding 6 — Ultrarare disorder with variable expressivity, adult-onset progression, and founder alleles

Finding 7 — Mechanism: dolichol deficiency impairs multiple ER glycosylation pathways

Finding 8 — CDG therapy is largely symptomatic; no established SRD5A3-specific therapy and no isolated mouse disease model

Detailed Section-by-Section Report

1. Disease Information

2. Etiology

3. Phenotypes

4. Genetic / Molecular Information

5. Environmental Information

6. Mechanism / Pathophysiology

7. Anatomical Structures Affected

8. Temporal Development

9. Inheritance and Population

10. Diagnostics

11. Outcome / Prognosis

12. Treatment

13. Prevention

14. Other Species / Natural Disease

15. Model Organisms

Mechanistic Model / Interpretation

Evidence Base

Limitations and Knowledge Gaps

Proposed Follow-up Experiments / Actions

Artifacts

Reference Validation

Term Validation

Terms the report names something else

Obsolete terms

Terms whose name is worth a second look

Terms named inconsistently

Prefixes with no resolver